draft-ietf-quic-transport-15.txt   draft-ietf-quic-transport-16.txt 
QUIC J. Iyengar, Ed. QUIC J. Iyengar, Ed.
Internet-Draft Fastly Internet-Draft Fastly
Intended status: Standards Track M. Thomson, Ed. Intended status: Standards Track M. Thomson, Ed.
Expires: April 6, 2019 Mozilla Expires: April 26, 2019 Mozilla
October 03, 2018 October 23, 2018
QUIC: A UDP-Based Multiplexed and Secure Transport QUIC: A UDP-Based Multiplexed and Secure Transport
draft-ietf-quic-transport-15 draft-ietf-quic-transport-16
Abstract Abstract
This document defines the core of the QUIC transport protocol. This This document defines the core of the QUIC transport protocol. This
document describes connection establishment, packet format, document describes connection establishment, packet format,
multiplexing, and reliability. Accompanying documents describe the multiplexing, and reliability. Accompanying documents describe the
cryptographic handshake and loss detection. cryptographic handshake and loss detection.
Note to Readers Note to Readers
skipping to change at page 1, line 44 skipping to change at page 1, line 44
Internet-Drafts are working documents of the Internet Engineering Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet- working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/. Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress." material or to cite them other than as "work in progress."
This Internet-Draft will expire on April 6, 2019. This Internet-Draft will expire on April 26, 2019.
Copyright Notice Copyright Notice
Copyright (c) 2018 IETF Trust and the persons identified as the Copyright (c) 2018 IETF Trust and the persons identified as the
document authors. All rights reserved. document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents Provisions Relating to IETF Documents
(https://trustee.ietf.org/license-info) in effect on the date of (https://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License. described in the Simplified BSD License.
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 5 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 5
2. Conventions and Definitions . . . . . . . . . . . . . . . . . 6 1.1. Document Structure . . . . . . . . . . . . . . . . . . . 6
2.1. Notational Conventions . . . . . . . . . . . . . . . . . 7 1.2. Conventions and Definitions . . . . . . . . . . . . . . . 7
3. Versions . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.3. Notational Conventions . . . . . . . . . . . . . . . . . 8
4. Packet Types and Formats . . . . . . . . . . . . . . . . . . 8 2. Streams . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.1. Long Header . . . . . . . . . . . . . . . . . . . . . . . 8 2.1. Stream Identifiers . . . . . . . . . . . . . . . . . . . 9
4.2. Short Header . . . . . . . . . . . . . . . . . . . . . . 11 2.2. Stream Concurrency . . . . . . . . . . . . . . . . . . . 10
4.3. Version Negotiation Packet . . . . . . . . . . . . . . . 12 2.3. Sending and Receiving Data . . . . . . . . . . . . . . . 11
4.4. Retry Packet . . . . . . . . . . . . . . . . . . . . . . 14 2.4. Stream Prioritization . . . . . . . . . . . . . . . . . . 11
4.5. Cryptographic Handshake Packets . . . . . . . . . . . . . 16 3. Stream States: Life of a Stream . . . . . . . . . . . . . . . 12
4.6. Initial Packet . . . . . . . . . . . . . . . . . . . . . 17 3.1. Send Stream States . . . . . . . . . . . . . . . . . . . 13
4.6.1. Connection IDs . . . . . . . . . . . . . . . . . . . 18 3.2. Receive Stream States . . . . . . . . . . . . . . . . . . 15
4.6.2. Tokens . . . . . . . . . . . . . . . . . . . . . . . 19 3.3. Permitted Frame Types . . . . . . . . . . . . . . . . . . 18
4.6.3. Starting Packet Numbers . . . . . . . . . . . . . . . 20 3.4. Bidirectional Stream States . . . . . . . . . . . . . . . 18
4.6.4. 0-RTT Packet Numbers . . . . . . . . . . . . . . . . 20 3.5. Solicited State Transitions . . . . . . . . . . . . . . . 19
4.6.5. Minimum Packet Size . . . . . . . . . . . . . . . . . 21 4. Flow Control . . . . . . . . . . . . . . . . . . . . . . . . 20
4.7. Handshake Packet . . . . . . . . . . . . . . . . . . . . 21 4.1. Handling of Stream Cancellation . . . . . . . . . . . . . 21
4.8. Protected Packets . . . . . . . . . . . . . . . . . . . . 22 4.2. Data Limit Increments . . . . . . . . . . . . . . . . . . 22
4.9. Coalescing Packets . . . . . . . . . . . . . . . . . . . 22 4.3. Stream Final Offset . . . . . . . . . . . . . . . . . . . 23
4.10. Connection ID Encoding . . . . . . . . . . . . . . . . . 23 4.4. Flow Control for Cryptographic Handshake . . . . . . . . 24
4.11. Packet Numbers . . . . . . . . . . . . . . . . . . . . . 24 4.5. Stream Limit Increment . . . . . . . . . . . . . . . . . 24
5. Frames and Frame Types . . . . . . . . . . . . . . . . . . . 27 5. Connections . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.1. Extension Frames . . . . . . . . . . . . . . . . . . . . 30 5.1. Connection ID . . . . . . . . . . . . . . . . . . . . . . 24
6. Life of a Connection . . . . . . . . . . . . . . . . . . . . 30 5.1.1. Issuing Connection IDs . . . . . . . . . . . . . . . 25
6.1. Connection ID . . . . . . . . . . . . . . . . . . . . . . 31 5.1.2. Consuming and Retiring Connection IDs . . . . . . . . 26
6.1.1. Issuing Connection IDs . . . . . . . . . . . . . . . 31 5.2. Matching Packets to Connections . . . . . . . . . . . . . 27
6.1.2. Consuming and Retiring Connection IDs . . . . . . . . 32 5.2.1. Client Packet Handling . . . . . . . . . . . . . . . 27
6.2. Matching Packets to Connections . . . . . . . . . . . . . 32 5.2.2. Server Packet Handling . . . . . . . . . . . . . . . 27
6.2.1. Client Packet Handling . . . . . . . . . . . . . . . 33 5.3. Life of a QUIC Connection . . . . . . . . . . . . . . . . 28
6.2.2. Server Packet Handling . . . . . . . . . . . . . . . 33 6. Version Negotiation . . . . . . . . . . . . . . . . . . . . . 28
6.3. Version Negotiation . . . . . . . . . . . . . . . . . . . 34 6.1. Sending Version Negotiation Packets . . . . . . . . . . . 29
6.3.1. Sending Version Negotiation Packets . . . . . . . . . 34 6.2. Handling Version Negotiation Packets . . . . . . . . . . 29
6.3.2. Handling Version Negotiation Packets . . . . . . . . 35 6.3. Using Reserved Versions . . . . . . . . . . . . . . . . . 30
6.3.3. Using Reserved Versions . . . . . . . . . . . . . . . 35 7. Cryptographic and Transport Handshake . . . . . . . . . . . . 31
6.4. Cryptographic and Transport Handshake . . . . . . . . . . 36 7.1. Example Handshake Flows . . . . . . . . . . . . . . . . . 32
6.5. Example Handshake Flows . . . . . . . . . . . . . . . . . 37 7.2. Negotiating Connection IDs . . . . . . . . . . . . . . . 33
6.6. Transport Parameters . . . . . . . . . . . . . . . . . . 38 7.3. Transport Parameters . . . . . . . . . . . . . . . . . . 34
6.6.1. Transport Parameter Definitions . . . . . . . . . . . 41 7.3.1. Values of Transport Parameters for 0-RTT . . . . . . 35
6.6.2. Values of Transport Parameters for 0-RTT . . . . . . 43 7.3.2. New Transport Parameters . . . . . . . . . . . . . . 36
6.6.3. New Transport Parameters . . . . . . . . . . . . . . 44 7.3.3. Version Negotiation Validation . . . . . . . . . . . 36
6.6.4. Version Negotiation Validation . . . . . . . . . . . 45 8. Address Validation . . . . . . . . . . . . . . . . . . . . . 37
6.7. Stateless Retries . . . . . . . . . . . . . . . . . . . . 46 8.1. Address Validation During Connection Establishment . . . 38
6.8. Using Explicit Congestion Notification . . . . . . . . . 46 8.1.1. Address Validation using Retry Packets . . . . . . . 38
6.9. Proof of Source Address Ownership . . . . . . . . . . . . 48 8.1.2. Address Validation for Future Connections . . . . . . 39
6.9.1. Client Address Validation Procedure . . . . . . . . . 49 8.1.3. Address Validation Token Integrity . . . . . . . . . 41
6.9.2. Address Validation for Future Connections . . . . . . 50 8.2. Path Validation . . . . . . . . . . . . . . . . . . . . . 41
6.9.3. Address Validation Token Integrity . . . . . . . . . 50 8.3. Initiating Path Validation . . . . . . . . . . . . . . . 42
6.10. Path Validation . . . . . . . . . . . . . . . . . . . . . 51 8.4. Path Validation Responses . . . . . . . . . . . . . . . . 42
6.10.1. Initiation . . . . . . . . . . . . . . . . . . . . . 51 8.5. Successful Path Validation . . . . . . . . . . . . . . . 42
6.10.2. Response . . . . . . . . . . . . . . . . . . . . . . 52 8.6. Failed Path Validation . . . . . . . . . . . . . . . . . 43
6.10.3. Completion . . . . . . . . . . . . . . . . . . . . . 52 9. Connection Migration . . . . . . . . . . . . . . . . . . . . 43
6.10.4. Abandonment . . . . . . . . . . . . . . . . . . . . 53 9.1. Probing a New Path . . . . . . . . . . . . . . . . . . . 44
6.11. Connection Migration . . . . . . . . . . . . . . . . . . 53 9.2. Initiating Connection Migration . . . . . . . . . . . . . 45
6.11.1. Probing a New Path . . . . . . . . . . . . . . . . . 54 9.3. Responding to Connection Migration . . . . . . . . . . . 45
6.11.2. Initiating Connection Migration . . . . . . . . . . 54 9.3.1. Handling Address Spoofing by a Peer . . . . . . . . . 46
6.11.3. Responding to Connection Migration . . . . . . . . . 55 9.3.2. Handling Address Spoofing by an On-path Attacker . . 46
6.11.4. Loss Detection and Congestion Control . . . . . . . 56 9.4. Loss Detection and Congestion Control . . . . . . . . . . 47
6.11.5. Privacy Implications of Connection Migration . . . . 57 9.5. Privacy Implications of Connection Migration . . . . . . 48
6.12. Server's Preferred Address . . . . . . . . . . . . . . . 58 9.6. Server's Preferred Address . . . . . . . . . . . . . . . 49
6.12.1. Communicating A Preferred Address . . . . . . . . . 59 9.6.1. Communicating A Preferred Address . . . . . . . . . . 49
6.12.2. Responding to Connection Migration . . . . . . . . . 59 9.6.2. Responding to Connection Migration . . . . . . . . . 49
6.12.3. Interaction of Client Migration and Preferred 9.6.3. Interaction of Client Migration and Preferred Address 50
Address . . . . . . . . . . . . . . . . . . . . . . 59 10. Connection Termination . . . . . . . . . . . . . . . . . . . 50
6.13. Connection Termination . . . . . . . . . . . . . . . . . 60 10.1. Closing and Draining Connection States . . . . . . . . . 51
6.13.1. Closing and Draining Connection States . . . . . . . 60 10.2. Idle Timeout . . . . . . . . . . . . . . . . . . . . . . 52
6.13.2. Idle Timeout . . . . . . . . . . . . . . . . . . . . 61 10.3. Immediate Close . . . . . . . . . . . . . . . . . . . . 52
6.13.3. Immediate Close . . . . . . . . . . . . . . . . . . 62 10.4. Stateless Reset . . . . . . . . . . . . . . . . . . . . 53
6.13.4. Stateless Reset . . . . . . . . . . . . . . . . . . 63 10.4.1. Detecting a Stateless Reset . . . . . . . . . . . . 56
7. Frame Types and Formats . . . . . . . . . . . . . . . . . . . 67 10.4.2. Calculating a Stateless Reset Token . . . . . . . . 56
7.1. Variable-Length Integer Encoding . . . . . . . . . . . . 67 10.4.3. Looping . . . . . . . . . . . . . . . . . . . . . . 57
7.2. PADDING Frame . . . . . . . . . . . . . . . . . . . . . . 68 11. Error Handling . . . . . . . . . . . . . . . . . . . . . . . 58
7.3. RST_STREAM Frame . . . . . . . . . . . . . . . . . . . . 68 11.1. Connection Errors . . . . . . . . . . . . . . . . . . . 58
7.4. CONNECTION_CLOSE frame . . . . . . . . . . . . . . . . . 69 11.2. Stream Errors . . . . . . . . . . . . . . . . . . . . . 59
7.5. APPLICATION_CLOSE frame . . . . . . . . . . . . . . . . . 70 12. Packets and Frames . . . . . . . . . . . . . . . . . . . . . 59
7.6. MAX_DATA Frame . . . . . . . . . . . . . . . . . . . . . 71 12.1. Protected Packets . . . . . . . . . . . . . . . . . . . 59
7.7. MAX_STREAM_DATA Frame . . . . . . . . . . . . . . . . . . 72 12.2. Coalescing Packets . . . . . . . . . . . . . . . . . . . 60
7.8. MAX_STREAM_ID Frame . . . . . . . . . . . . . . . . . . . 73 12.3. Packet Numbers . . . . . . . . . . . . . . . . . . . . . 61
7.9. PING Frame . . . . . . . . . . . . . . . . . . . . . . . 73 12.4. Frames and Frame Types . . . . . . . . . . . . . . . . . 62
7.10. BLOCKED Frame . . . . . . . . . . . . . . . . . . . . . . 74 13. Packetization and Reliability . . . . . . . . . . . . . . . . 65
7.11. STREAM_BLOCKED Frame . . . . . . . . . . . . . . . . . . 74 13.1. Packet Processing and Acknowledgment . . . . . . . . . . 66
7.12. STREAM_ID_BLOCKED Frame . . . . . . . . . . . . . . . . . 75 13.1.1. Sending ACK Frames . . . . . . . . . . . . . . . . . 66
7.13. NEW_CONNECTION_ID Frame . . . . . . . . . . . . . . . . . 75 13.1.2. ACK Frames and Packet Protection . . . . . . . . . . 67
7.14. RETIRE_CONNECTION_ID Frame . . . . . . . . . . . . . . . 77 13.2. Retransmission of Information . . . . . . . . . . . . . 67
7.15. STOP_SENDING Frame . . . . . . . . . . . . . . . . . . . 77 13.3. Explicit Congestion Notification . . . . . . . . . . . . 69
7.16. ACK Frame . . . . . . . . . . . . . . . . . . . . . . . . 78 13.3.1. ECN Counters . . . . . . . . . . . . . . . . . . . . 70
7.16.1. ACK Block Section . . . . . . . . . . . . . . . . . 79 13.3.2. ECN Verification . . . . . . . . . . . . . . . . . . 70
7.16.2. ECN section . . . . . . . . . . . . . . . . . . . . 81 14. Packet Size . . . . . . . . . . . . . . . . . . . . . . . . . 71
7.16.3. Sending ACK Frames . . . . . . . . . . . . . . . . . 82 14.1. Path Maximum Transmission Unit . . . . . . . . . . . . . 72
7.16.4. ACK Frames and Packet Protection . . . . . . . . . . 83 14.1.1. IPv4 PMTU Discovery . . . . . . . . . . . . . . . . 73
7.17. PATH_CHALLENGE Frame . . . . . . . . . . . . . . . . . . 83 14.2. Special Considerations for Packetization Layer PMTU
7.18. PATH_RESPONSE Frame . . . . . . . . . . . . . . . . . . . 84 Discovery . . . . . . . . . . . . . . . . . . . . . . . 73
7.19. NEW_TOKEN frame . . . . . . . . . . . . . . . . . . . . . 84 15. Versions . . . . . . . . . . . . . . . . . . . . . . . . . . 74
7.20. STREAM Frames . . . . . . . . . . . . . . . . . . . . . . 84 16. Variable-Length Integer Encoding . . . . . . . . . . . . . . 75
7.21. CRYPTO Frame . . . . . . . . . . . . . . . . . . . . . . 86 17. Packet Formats . . . . . . . . . . . . . . . . . . . . . . . 75
8. Packetization and Reliability . . . . . . . . . . . . . . . . 87 17.1. Packet Number Encoding and Decoding . . . . . . . . . . 76
8.1. Packet Processing and Acknowledgment . . . . . . . . . . 87 17.2. Long Header Packet . . . . . . . . . . . . . . . . . . . 77
8.2. Retransmission of Information . . . . . . . . . . . . . . 88 17.3. Short Header Packet . . . . . . . . . . . . . . . . . . 79
8.3. Packet Size . . . . . . . . . . . . . . . . . . . . . . . 90 17.4. Version Negotiation Packet . . . . . . . . . . . . . . . 81
8.4. Path Maximum Transmission Unit . . . . . . . . . . . . . 90 17.5. Initial Packet . . . . . . . . . . . . . . . . . . . . . 82
8.4.1. IPv4 PMTU Discovery . . . . . . . . . . . . . . . . . 91 17.5.1. Starting Packet Numbers . . . . . . . . . . . . . . 84
8.4.2. Special Considerations for Packetization Layer PMTU 17.5.2. 0-RTT Packet Numbers . . . . . . . . . . . . . . . . 84
Discovery . . . . . . . . . . . . . . . . . . . . . . 92 17.6. Handshake Packet . . . . . . . . . . . . . . . . . . . . 85
9. Streams: QUIC's Data Structuring Abstraction . . . . . . . . 92 17.7. Retry Packet . . . . . . . . . . . . . . . . . . . . . . 85
9.1. Stream Identifiers . . . . . . . . . . . . . . . . . . . 93 18. Transport Parameter Encoding . . . . . . . . . . . . . . . . 88
9.2. Stream States . . . . . . . . . . . . . . . . . . . . . . 94 18.1. Transport Parameter Definitions . . . . . . . . . . . . 90
9.2.1. Send Stream States . . . . . . . . . . . . . . . . . 95 19. Frame Types and Formats . . . . . . . . . . . . . . . . . . . 92
9.2.2. Receive Stream States . . . . . . . . . . . . . . . . 97 19.1. PADDING Frame . . . . . . . . . . . . . . . . . . . . . 93
9.2.3. Permitted Frame Types . . . . . . . . . . . . . . . . 99 19.2. RST_STREAM Frame . . . . . . . . . . . . . . . . . . . . 93
9.2.4. Bidirectional Stream States . . . . . . . . . . . . . 99 19.3. CONNECTION_CLOSE frame . . . . . . . . . . . . . . . . . 94
9.3. Solicited State Transitions . . . . . . . . . . . . . . . 101 19.4. APPLICATION_CLOSE frame . . . . . . . . . . . . . . . . 95
9.4. Stream Concurrency . . . . . . . . . . . . . . . . . . . 101 19.5. MAX_DATA Frame . . . . . . . . . . . . . . . . . . . . . 95
9.5. Sending and Receiving Data . . . . . . . . . . . . . . . 102 19.6. MAX_STREAM_DATA Frame . . . . . . . . . . . . . . . . . 96
9.6. Stream Prioritization . . . . . . . . . . . . . . . . . . 102 19.7. MAX_STREAM_ID Frame . . . . . . . . . . . . . . . . . . 97
10. Flow Control . . . . . . . . . . . . . . . . . . . . . . . . 103 19.8. PING Frame . . . . . . . . . . . . . . . . . . . . . . . 98
10.1. Edge Cases and Other Considerations . . . . . . . . . . 105 19.9. BLOCKED Frame . . . . . . . . . . . . . . . . . . . . . 98
10.1.1. Response to a RST_STREAM . . . . . . . . . . . . . . 105 19.10. STREAM_BLOCKED Frame . . . . . . . . . . . . . . . . . . 99
10.1.2. Data Limit Increments . . . . . . . . . . . . . . . 105 19.11. STREAM_ID_BLOCKED Frame . . . . . . . . . . . . . . . . 99
10.2. Stream Limit Increment . . . . . . . . . . . . . . . . . 106 19.12. NEW_CONNECTION_ID Frame . . . . . . . . . . . . . . . . 100
10.2.1. Blocking on Flow Control . . . . . . . . . . . . . . 106 19.13. RETIRE_CONNECTION_ID Frame . . . . . . . . . . . . . . . 101
10.3. Stream Final Offset . . . . . . . . . . . . . . . . . . 107 19.14. STOP_SENDING Frame . . . . . . . . . . . . . . . . . . . 102
10.4. Flow Control for Cryptographic Handshake . . . . . . . . 107 19.15. ACK Frame . . . . . . . . . . . . . . . . . . . . . . . 102
11. Error Handling . . . . . . . . . . . . . . . . . . . . . . . 107 19.15.1. ACK Block Section . . . . . . . . . . . . . . . . . 104
11.1. Connection Errors . . . . . . . . . . . . . . . . . . . 108 19.15.2. ECN section . . . . . . . . . . . . . . . . . . . . 105
11.2. Stream Errors . . . . . . . . . . . . . . . . . . . . . 108 19.16. PATH_CHALLENGE Frame . . . . . . . . . . . . . . . . . . 106
11.3. Transport Error Codes . . . . . . . . . . . . . . . . . 109 19.17. PATH_RESPONSE Frame . . . . . . . . . . . . . . . . . . 107
11.4. Application Protocol Error Codes . . . . . . . . . . . . 110 19.18. NEW_TOKEN frame . . . . . . . . . . . . . . . . . . . . 107
12. Security Considerations . . . . . . . . . . . . . . . . . . . 110 19.19. STREAM Frames . . . . . . . . . . . . . . . . . . . . . 107
12.1. Handshake Denial of Service . . . . . . . . . . . . . . 110 19.20. CRYPTO Frame . . . . . . . . . . . . . . . . . . . . . . 109
12.2. Spoofed ACK Attack . . . . . . . . . . . . . . . . . . . 111 19.21. Extension Frames . . . . . . . . . . . . . . . . . . . . 110
12.3. Optimistic ACK Attack . . . . . . . . . . . . . . . . . 112 20. Transport Error Codes . . . . . . . . . . . . . . . . . . . . 110
12.4. Slowloris Attacks . . . . . . . . . . . . . . . . . . . 112 20.1. Application Protocol Error Codes . . . . . . . . . . . . 111
12.5. Stream Fragmentation and Reassembly Attacks . . . . . . 113 21. Security Considerations . . . . . . . . . . . . . . . . . . . 112
12.6. Stream Commitment Attack . . . . . . . . . . . . . . . . 113 21.1. Handshake Denial of Service . . . . . . . . . . . . . . 112
12.7. Explicit Congestion Notification Attacks . . . . . . . . 114 21.2. Spoofed ACK Attack . . . . . . . . . . . . . . . . . . . 113
12.8. Stateless Reset Oracle . . . . . . . . . . . . . . . . . 114 21.3. Optimistic ACK Attack . . . . . . . . . . . . . . . . . 113
13. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 114 21.4. Slowloris Attacks . . . . . . . . . . . . . . . . . . . 114
13.1. QUIC Transport Parameter Registry . . . . . . . . . . . 114 21.5. Stream Fragmentation and Reassembly Attacks . . . . . . 114
13.2. QUIC Frame Type Registry . . . . . . . . . . . . . . . . 116 21.6. Stream Commitment Attack . . . . . . . . . . . . . . . . 114
13.3. QUIC Transport Error Codes Registry . . . . . . . . . . 117 21.7. Explicit Congestion Notification Attacks . . . . . . . . 115
14. References . . . . . . . . . . . . . . . . . . . . . . . . . 120 21.8. Stateless Reset Oracle . . . . . . . . . . . . . . . . . 115
14.1. Normative References . . . . . . . . . . . . . . . . . . 120 22. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 116
14.2. Informative References . . . . . . . . . . . . . . . . . 121 22.1. QUIC Transport Parameter Registry . . . . . . . . . . . 116
Appendix A. Sample Packet Number Decoding Algorithm . . . . . . 122 22.2. QUIC Frame Type Registry . . . . . . . . . . . . . . . . 117
Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 123 22.3. QUIC Transport Error Codes Registry . . . . . . . . . . 118
B.1. Since draft-ietf-quic-transport-14 . . . . . . . . . . . 123 23. References . . . . . . . . . . . . . . . . . . . . . . . . . 121
B.2. Since draft-ietf-quic-transport-13 . . . . . . . . . . . 124 23.1. Normative References . . . . . . . . . . . . . . . . . . 121
B.3. Since draft-ietf-quic-transport-12 . . . . . . . . . . . 124 23.2. Informative References . . . . . . . . . . . . . . . . . 122
B.4. Since draft-ietf-quic-transport-11 . . . . . . . . . . . 125 Appendix A. Sample Packet Number Decoding Algorithm . . . . . . 123
B.5. Since draft-ietf-quic-transport-10 . . . . . . . . . . . 126 Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 124
B.6. Since draft-ietf-quic-transport-09 . . . . . . . . . . . 126 B.1. Since draft-ietf-quic-transport-15 . . . . . . . . . . . 124
B.7. Since draft-ietf-quic-transport-08 . . . . . . . . . . . 127 B.2. Since draft-ietf-quic-transport-14 . . . . . . . . . . . 124
B.8. Since draft-ietf-quic-transport-07 . . . . . . . . . . . 127 B.3. Since draft-ietf-quic-transport-13 . . . . . . . . . . . 125
B.9. Since draft-ietf-quic-transport-06 . . . . . . . . . . . 128 B.4. Since draft-ietf-quic-transport-12 . . . . . . . . . . . 126
B.10. Since draft-ietf-quic-transport-05 . . . . . . . . . . . 129 B.5. Since draft-ietf-quic-transport-11 . . . . . . . . . . . 126
B.11. Since draft-ietf-quic-transport-04 . . . . . . . . . . . 129 B.6. Since draft-ietf-quic-transport-10 . . . . . . . . . . . 127
B.12. Since draft-ietf-quic-transport-03 . . . . . . . . . . . 130 B.7. Since draft-ietf-quic-transport-09 . . . . . . . . . . . 127
B.13. Since draft-ietf-quic-transport-02 . . . . . . . . . . . 130 B.8. Since draft-ietf-quic-transport-08 . . . . . . . . . . . 128
B.14. Since draft-ietf-quic-transport-01 . . . . . . . . . . . 131 B.9. Since draft-ietf-quic-transport-07 . . . . . . . . . . . 129
B.15. Since draft-ietf-quic-transport-00 . . . . . . . . . . . 133 B.10. Since draft-ietf-quic-transport-06 . . . . . . . . . . . 130
B.16. Since draft-hamilton-quic-transport-protocol-01 . . . . . 133 B.11. Since draft-ietf-quic-transport-05 . . . . . . . . . . . 130
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 133 B.12. Since draft-ietf-quic-transport-04 . . . . . . . . . . . 130
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . 134 B.13. Since draft-ietf-quic-transport-03 . . . . . . . . . . . 131
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 134 B.14. Since draft-ietf-quic-transport-02 . . . . . . . . . . . 131
B.15. Since draft-ietf-quic-transport-01 . . . . . . . . . . . 132
B.16. Since draft-ietf-quic-transport-00 . . . . . . . . . . . 134
B.17. Since draft-hamilton-quic-transport-protocol-01 . . . . . 134
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 134
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 135
1. Introduction 1. Introduction
QUIC is a multiplexed and secure transport protocol that runs on top QUIC is a multiplexed and secure transport protocol that runs on top
of UDP. QUIC aims to provide a flexible set of features that allow of UDP. QUIC aims to provide a flexible set of features that allow
it to be a general-purpose secure transport for multiple it to be a general-purpose secure transport for multiple
applications. applications.
o Version negotiation o Version negotiation
skipping to change at page 6, line 18 skipping to change at page 6, line 25
o Stream and connection-level flow control o Stream and connection-level flow control
o Connection migration and resilience to NAT rebinding o Connection migration and resilience to NAT rebinding
QUIC uses UDP as a substrate to avoid requiring changes in legacy QUIC uses UDP as a substrate to avoid requiring changes in legacy
client operating systems and middleboxes. QUIC authenticates all of client operating systems and middleboxes. QUIC authenticates all of
its headers and encrypts most of the data it exchanges, including its its headers and encrypts most of the data it exchanges, including its
signaling. This allows the protocol to evolve without incurring a signaling. This allows the protocol to evolve without incurring a
dependency on upgrades to middleboxes. dependency on upgrades to middleboxes.
This document describes the core QUIC protocol, including the 1.1. Document Structure
conceptual design, wire format, and mechanisms of the QUIC protocol
for connection establishment, stream multiplexing, stream and This document describes the core QUIC protocol, and is structured as
connection-level flow control, connection migration, and data follows:
reliability.
o Streams are the basic service abstraction that QUIC provides.
* Section 2 describes core concepts related to streams,
* Section 3 provides a reference model for stream states, and
* Section 4 outlines the operation of flow control.
o Connections are the context in which QUIC endpoints communicate.
* Section 5 describes core concepts related to connections,
* Section 6 describes version negotiation,
* Section 7 details the process for establishing connections,
* Section 8 specifies critical denial of service mitigation
mechanisms,
* Section 9 describes how endpoints migrate a connection to use a
new network paths, and
* Section 10 lists the options for terminating an open
connection.
o Packets and frames are the basic unit used by QUIC to communicate.
* Section 12 describes concepts related to packets and frames,
* Section 13 defines models for the transmission, retransmission,
and acknowledgement of information, and
* Section 14 contains a rules for managing the size of packets.
o Details of encoding of QUIC protocol elements is described in:
* Section 15 (Versions),
* Section 17 (Packet Headers),
* Section 18 (Transport Parameters),
* Section 19 (Frames), and
* Section 20 (Errors).
Accompanying documents describe QUIC's loss detection and congestion Accompanying documents describe QUIC's loss detection and congestion
control [QUIC-RECOVERY], and the use of TLS 1.3 for key negotiation control [QUIC-RECOVERY], and the use of TLS 1.3 for key negotiation
[QUIC-TLS]. [QUIC-TLS].
QUIC version 1 conforms to the protocol invariants in QUIC version 1 conforms to the protocol invariants in
[QUIC-INVARIANTS]. [QUIC-INVARIANTS].
2. Conventions and Definitions 1.2. Conventions and Definitions
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in BCP "OPTIONAL" in this document are to be interpreted as described in BCP
14 [RFC2119] [RFC8174] when, and only when, they appear in all 14 [RFC2119] [RFC8174] when, and only when, they appear in all
capitals, as shown here. capitals, as shown here.
Definitions of terms that are used in this document: Definitions of terms that are used in this document:
Client: The endpoint initiating a QUIC connection. Client: The endpoint initiating a QUIC connection.
Server: The endpoint accepting incoming QUIC connections. Server: The endpoint accepting incoming QUIC connections.
Endpoint: The client or server end of a connection. Endpoint: The client or server end of a connection.
Stream: A logical, bi-directional channel of ordered bytes within a Stream: A logical unidirectional or bidirectional channel of ordered
QUIC connection. bytes within a QUIC connection.
Connection: A conversation between two QUIC endpoints with a single Connection: A conversation between two QUIC endpoints with a single
encryption context that multiplexes streams within it. encryption context that multiplexes streams within it.
Connection ID: An opaque identifier that is used to identify a QUIC Connection ID: An opaque identifier that is used to identify a QUIC
connection at an endpoint. Each endpoint sets a value that its connection at an endpoint. Each endpoint sets a value that its
peer includes in packets. peer includes in packets.
QUIC packet: The smallest unit of data that can be exchanged by QUIC QUIC packet: The smallest unit of data that can be exchanged by QUIC
endpoints. endpoints.
QUIC is a name, not an acronym. QUIC is a name, not an acronym.
2.1. Notational Conventions 1.3. Notational Conventions
Packet and frame diagrams use the format described in Section 3.1 of Packet and frame diagrams use the format described in Section 3.1 of
[RFC2360], with the following additional conventions: [RFC2360], with the following additional conventions:
[x] Indicates that x is optional [x] Indicates that x is optional
x (A) Indicates that x is A bits long x (A) Indicates that x is A bits long
x (A/B/C) ... Indicates that x is one of A, B, or C bits long x (A/B/C) ... Indicates that x is one of A, B, or C bits long
x (i) ... Indicates that x uses the variable-length encoding in x (i) ... Indicates that x uses the variable-length encoding in
Section 7.1 Section 16
x (*) ... Indicates that x is variable-length x (*) ... Indicates that x is variable-length
3. Versions 2. Streams
QUIC versions are identified using a 32-bit unsigned number.
The version 0x00000000 is reserved to represent version negotiation.
This version of the specification is identified by the number
0x00000001.
Other versions of QUIC might have different properties to this
version. The properties of QUIC that are guaranteed to be consistent
across all versions of the protocol are described in
[QUIC-INVARIANTS].
Version 0x00000001 of QUIC uses TLS as a cryptographic handshake
protocol, as described in [QUIC-TLS].
Versions with the most significant 16 bits of the version number
cleared are reserved for use in future IETF consensus documents.
Versions that follow the pattern 0x?a?a?a?a are reserved for use in
forcing version negotiation to be exercised. That is, any version
number where the low four bits of all octets is 1010 (in binary). A
client or server MAY advertise support for any of these reserved
versions.
Reserved version numbers will probably never represent a real
protocol; a client MAY use one of these version numbers with the
expectation that the server will initiate version negotiation; a
server MAY advertise support for one of these versions and can expect
that clients ignore the value.
[[RFC editor: please remove the remainder of this section before
publication.]]
The version number for the final version of this specification
(0x00000001), is reserved for the version of the protocol that is
published as an RFC.
Version numbers used to identify IETF drafts are created by adding
the draft number to 0xff000000. For example, draft-ietf-quic-
transport-13 would be identified as 0xff00000D.
Implementors are encouraged to register version numbers of QUIC that
they are using for private experimentation on the GitHub wiki at
<https://github.com/quicwg/base-drafts/wiki/QUIC-Versions>.
4. Packet Types and Formats
We first describe QUIC's packet types and their formats, since some
are referenced in subsequent mechanisms.
All numeric values are encoded in network byte order (that is, big-
endian) and all field sizes are in bits. When discussing individual
bits of fields, the least significant bit is referred to as bit 0.
Hexadecimal notation is used for describing the value of fields.
Any QUIC packet has either a long or a short header, as indicated by
the Header Form bit. Long headers are expected to be used early in
the connection before version negotiation and establishment of 1-RTT
keys. Short headers are minimal version-specific headers, which are
used after version negotiation and 1-RTT keys are established.
4.1. Long Header
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+
|1| Type (7) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Version (32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|DCIL(4)|SCIL(4)|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Source Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Length (i) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Packet Number (8/16/32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Payload (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 1: Long Header Packet Format
Long headers are used for packets that are sent prior to the
completion of version negotiation and establishment of 1-RTT keys.
Once both conditions are met, a sender switches to sending packets
using the short header (Section 4.2). The long form allows for
special packets - such as the Version Negotiation packet - to be
represented in this uniform fixed-length packet format. Packets that
use the long header contain the following fields:
Header Form: The most significant bit (0x80) of octet 0 (the first
octet) is set to 1 for long headers.
Long Packet Type: The remaining seven bits of octet 0 contain the
packet type. This field can indicate one of 128 packet types.
The types specified for this version are listed in Table 1.
Version: The QUIC Version is a 32-bit field that follows the Type.
This field indicates which version of QUIC is in use and
determines how the rest of the protocol fields are interpreted.
DCIL and SCIL: The octet following the version contains the lengths
of the two connection ID fields that follow it. These lengths are
encoded as two 4-bit unsigned integers. The Destination
Connection ID Length (DCIL) field occupies the 4 high bits of the
octet and the Source Connection ID Length (SCIL) field occupies
the 4 low bits of the octet. An encoded length of 0 indicates
that the connection ID is also 0 octets in length. Non-zero
encoded lengths are increased by 3 to get the full length of the
connection ID, producing a length between 4 and 18 octets
inclusive. For example, an octet with the value 0x50 describes an
8-octet Destination Connection ID and a zero-length Source
Connection ID.
Destination Connection ID: The Destination Connection ID field
follows the connection ID lengths and is either 0 octets in length
or between 4 and 18 octets. Section 4.10 describes the use of
this field in more detail.
Source Connection ID: The Source Connection ID field follows the
Destination Connection ID and is either 0 octets in length or
between 4 and 18 octets. Section 4.10 describes the use of this
field in more detail.
Length: The length of the remainder of the packet (that is, the
Packet Number and Payload fields) in octets, encoded as a
variable-length integer (Section 7.1).
Packet Number: The packet number field is 1, 2, or 4 octets long.
The packet number has confidentiality protection separate from
packet protection, as described in Section 5.3 of [QUIC-TLS]. The
length of the packet number field is encoded in the plaintext
packet number. See Section 4.11 for details.
Payload: The payload of the packet.
The following packet types are defined:
+------+-----------------+-------------+
| Type | Name | Section |
+------+-----------------+-------------+
| 0x7F | Initial | Section 4.6 |
| | | |
| 0x7E | Retry | Section 4.4 |
| | | |
| 0x7D | Handshake | Section 4.7 |
| | | |
| 0x7C | 0-RTT Protected | Section 4.8 |
+------+-----------------+-------------+
Table 1: Long Header Packet Types
The header form, type, connection ID lengths octet, destination and
source connection IDs, and version fields of a long header packet are
version-independent. The packet number and values for packet types
defined in Table 1 are version-specific. See [QUIC-INVARIANTS] for
details on how packets from different versions of QUIC are
interpreted.
The interpretation of the fields and the payload are specific to a
version and packet type. Type-specific semantics for this version
are described in the following sections.
The end of the packet is determined by the Length field. The Length
field covers both the Packet Number and Payload fields, both of which
are confidentiality protected and initially of unknown length. The
size of the Payload field is learned once the packet number
protection is removed.
Senders can sometimes coalesce multiple packets into one UDP
datagram. See Section 4.9 for more details.
4.2. Short Header
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+
|0|K|1|1|0|R R R|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Packet Number (8/16/32) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Protected Payload (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 2: Short Header Packet Format
The short header can be used after the version and 1-RTT keys are
negotiated. Packets that use the short header contain the following
fields:
Header Form: The most significant bit (0x80) of octet 0 is set to 0
for the short header.
Key Phase Bit: The second bit (0x40) of octet 0 indicates the key
phase, which allows a recipient of a packet to identify the packet
protection keys that are used to protect the packet. See
[QUIC-TLS] for details.
[[Editor's Note: this section should be removed and the bit
definitions changed before this draft goes to the IESG.]]
Third Bit: The third bit (0x20) of octet 0 is set to 1.
[[Editor's Note: this section should be removed and the bit
definitions changed before this draft goes to the IESG.]]
Fourth Bit: The fourth bit (0x10) of octet 0 is set to 1.
[[Editor's Note: this section should be removed and the bit
definitions changed before this draft goes to the IESG.]]
Google QUIC Demultiplexing Bit: The fifth bit (0x8) of octet 0 is
set to 0. This allows implementations of Google QUIC to
distinguish Google QUIC packets from short header packets sent by
a client because Google QUIC servers expect the connection ID to
always be present. The special interpretation of this bit SHOULD
be removed from this specification when Google QUIC has finished
transitioning to the new header format.
Reserved: The sixth, seventh, and eighth bits (0x7) of octet 0 are
reserved for experimentation. Endpoints MUST ignore these bits on
packets they receive unless they are participating in an
experiment that uses these bits. An endpoint not actively using
these bits SHOULD set the value randomly on packets they send to
protect against unwanted inference about particular values.
Destination Connection ID: The Destination Connection ID is a
connection ID that is chosen by the intended recipient of the
packet. See Section 6.1 for more details.
Packet Number: The packet number field is 1, 2, or 4 octets long.
The packet number has confidentiality protection separate from
packet protection, as described in Section 5.3 of [QUIC-TLS]. The
length of the packet number field is encoded in the plaintext
packet number. See Section 4.11 for details.
Protected Payload: Packets with a short header always include a
1-RTT protected payload.
The header form and connection ID field of a short header packet are
version-independent. The remaining fields are specific to the
selected QUIC version. See [QUIC-INVARIANTS] for details on how
packets from different versions of QUIC are interpreted.
4.3. Version Negotiation Packet
A Version Negotiation packet is inherently not version-specific, and
does not use the long packet header (see Section 4.1. Upon receipt
by a client, it will appear to be a packet using the long header, but
will be identified as a Version Negotiation packet based on the
Version field having a value of 0.
The Version Negotiation packet is a response to a client packet that
contains a version that is not supported by the server, and is only
sent by servers.
The layout of a Version Negotiation packet is:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+
|1| Unused (7) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Version (32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|DCIL(4)|SCIL(4)|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Source Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Supported Version 1 (32) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| [Supported Version 2 (32)] ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| [Supported Version N (32)] ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 3: Version Negotiation Packet
The value in the Unused field is selected randomly by the server.
The Version field of a Version Negotiation packet MUST be set to
0x00000000.
The server MUST include the value from the Source Connection ID field
of the packet it receives in the Destination Connection ID field.
The value for Source Connection ID MUST be copied from the
Destination Connection ID of the received packet, which is initially
randomly selected by a client. Echoing both connection IDs gives
clients some assurance that the server received the packet and that
the Version Negotiation packet was not generated by an off-path
attacker.
The remainder of the Version Negotiation packet is a list of 32-bit
versions which the server supports.
A Version Negotiation packet cannot be explicitly acknowledged in an
ACK frame by a client. Receiving another Initial packet implicitly
acknowledges a Version Negotiation packet.
The Version Negotiation packet does not include the Packet Number and
Length fields present in other packets that use the long header form.
Consequently, a Version Negotiation packet consumes an entire UDP
datagram.
See Section 6.3 for a description of the version negotiation process.
4.4. Retry Packet
A Retry packet uses a long packet header with a type value of 0x7E.
It carries an address validation token created by the server. It is
used by a server that wishes to perform a stateless retry (see
Section 6.7).
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+
|1| 0x7e |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Version (32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|DCIL(4)|SCIL(4)|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Source Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ODCIL(8) | Original Destination Connection ID (*) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Retry Token (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 4: Retry Packet
A Retry packet (shown in Figure 4) only uses the invariant portion of
the long packet header [QUIC-INVARIANTS]; that is, the fields up to
and including the Destination and Source Connection ID fields. A
Retry packet does not contain any protected fields. Like Version
Negotiation, a Retry packet contains the long header including the
connection IDs, but omits the Length, Packet Number, and Payload
fields. These are replaced with:
ODCIL: The length of the Original Destination Connection ID field.
The length is encoded in the least significant 4 bits of the
octet, using the same encoding as the DCIL and SCIL fields. The
most significant 4 bits of this octet are reserved. Unless a use
for these bits has been negotiated, endpoints SHOULD send
randomized values and MUST ignore any value that it receives.
Original Destination Connection ID: The Original Destination
Connection ID contains the value of the Destination Connection ID
from the Initial packet that this Retry is in response to. The
length of this field is given in ODCIL.
Retry Token: An opaque token that the server can use to validate the
client's address.
The server populates the Destination Connection ID with the Streams in QUIC provide a lightweight, ordered byte-stream
connection ID that the client included in the Source Connection ID of abstraction.
the Initial packet.
The server includes a connection ID of its choice in the Source There are two basic types of stream in QUIC. Unidirectional streams
Connection ID field. This value MUST not be equal to the Destination carry data in one direction: from the initiator of the stream to its
Connection ID field of the packet sent by the client. The client peer; bidirectional streams allow for data to be sent in both
MUST use this connection ID in the Destination Connection ID of directions. Different stream identifiers are used to distinguish
subsequent packets that it sends. between unidirectional and bidirectional streams, as well as to
create a separation between streams that are initiated by the client
and server (see Section 2.1).
A server MAY send Retry packets in response to Initial and 0-RTT Either type of stream can be created by either endpoint, can
packets. A server can either discard or buffer 0-RTT packets that it concurrently send data interleaved with other streams, and can be
receives. A server can send multiple Retry packets as it receives cancelled.
Initial or 0-RTT packets.
A client MUST accept and process at most one Retry packet for each Streams can be created by sending data. Other processes associated
connection attempt. After the client has received and processed an with stream management - ending, cancelling, and managing flow
Initial or Retry packet from the server, it MUST discard any control - are all designed to impose minimal overheads. For
subsequent Retry packets that it receives. instance, a single STREAM frame (Section 19.19) can open, carry data
for, and close a stream. Streams can also be long-lived and can last
the entire duration of a connection.
Clients MUST discard Retry packets that contain an Original Stream offsets allow for the octets on a stream to be placed in
Destination Connection ID field that does not match the Destination order. An endpoint MUST be capable of delivering data received on a
Connection ID from its Initial packet. This prevents an off-path stream in order. Implementations MAY choose to offer the ability to
attacker from injecting a Retry packet. deliver data out of order. There is no means of ensuring ordering
between octets on different streams.
The client responds to a Retry packet with an Initial packet that Streams are individually flow controlled, allowing an endpoint to
includes the provided Retry Token to continue connection limit memory commitment and to apply back pressure. The creation of
establishment. streams is also flow controlled, with each peer declaring the maximum
stream ID it is willing to accept at a given time.
A client sets the Destination Connection ID field of this Initial An alternative view of QUIC streams is as an elastic "message"
packet to the value from the Source Connection ID in the Retry abstraction, similar to the way ephemeral streams are used in SST
packet. Changing Destination Connection ID also results in a change [SST], which may be a more appealing description for some
to the keys used to protect the Initial packet. It also sets the applications.
Token field to the token provided in the Retry. The client MUST NOT
change the Source Connection ID because the server could include the
connection ID as part of its token validation logic (see
Section 4.6.2).
All subsequent Initial packets from the client MUST use the 2.1. Stream Identifiers
connection ID and token values from the Retry packet. Aside from
this, the Initial packet sent by the client is subject to the same
restrictions as the first Initial packet. A client can either reuse
the cryptographic handshake message or construct a new one at its
discretion.
A client MAY attempt 0-RTT after receiving a Retry packet by sending Streams are identified by an unsigned 62-bit integer, referred to as
0-RTT packets to the connection ID provided by the server. A client the Stream ID. Stream IDs are encoded as a variable-length integer
that sends additional 0-RTT packets without constructing a new (see Section 16). The least significant two bits of the Stream ID
cryptographic handshake message MUST NOT reset the packet number to 0 are used to identify the type of stream (unidirectional or
after a Retry packet, see Section 4.6.4. bidirectional) and the initiator of the stream.
A server acknowledges the use of a Retry packet for a connection The least significant bit (0x1) of the Stream ID identifies the
using the original_connection_id transport parameter (see initiator of the stream. Clients initiate even-numbered streams
Section 6.6.1). If the server sends a Retry packet, it MUST include (those with the least significant bit set to 0); servers initiate
the value of the Original Destination Connection ID field of the odd-numbered streams (with the bit set to 1). Separation of the
Retry packet (that is, the Destination Connection ID field from the stream identifiers ensures that client and server are able to open
client's first Initial packet) in the transport parameter. streams without the latency imposed by negotiating for an identifier.
If the client received and processed a Retry packet, it validates If an endpoint receives a frame for a stream that it expects to
that the original_connection_id transport parameter is present and initiate (i.e., odd-numbered for the client or even-numbered for the
correct; otherwise, it validates that the transport parameter is server), but which it has not yet opened, it MUST close the
absent. A client MUST treat a failed validation as a connection connection with error code STREAM_STATE_ERROR.
error of type TRANSPORT_PARAMETER_ERROR.
A Retry packet does not include a packet number and cannot be The second least significant bit (0x2) of the Stream ID
explicitly acknowledged by a client. differentiates between unidirectional streams and bidirectional
streams. Unidirectional streams always have this bit set to 1 and
bidirectional streams have this bit set to 0.
4.5. Cryptographic Handshake Packets The two type bits from a Stream ID therefore identify streams as
summarized in Table 1.
Once version negotiation is complete, the cryptographic handshake is +----------+----------------------------------+
used to agree on cryptographic keys. The cryptographic handshake is | Low Bits | Stream Type |
carried in Initial (Section 4.6) and Handshake (Section 4.7) packets. +----------+----------------------------------+
| 0x0 | Client-Initiated, Bidirectional |
| | |
| 0x1 | Server-Initiated, Bidirectional |
| | |
| 0x2 | Client-Initiated, Unidirectional |
| | |
| 0x3 | Server-Initiated, Unidirectional |
+----------+----------------------------------+
All these packets use the long header and contain the current QUIC Table 1: Stream ID Types
version in the version field.
In order to prevent tampering by version-unaware middleboxes, Initial The first bidirectional stream opened by the client is stream 0.
packets are protected with connection- and version-specific keys
(Initial keys) as described in [QUIC-TLS]. This protection does not
provide confidentiality or integrity against on-path attackers, but
provides some level of protection against off-path attackers.
4.6. Initial Packet A QUIC endpoint MUST NOT reuse a Stream ID. Streams of each type are
created in numeric order. Streams that are used out of order result
in opening all lower-numbered streams of the same type in the same
direction.
The Initial packet uses long headers with a type value of 0x7F. It 2.2. Stream Concurrency
carries the first CRYPTO frames sent by the client and server to
perform key exchange, and carries ACKs in either direction. The
Initial packet is protected by Initial keys as described in
[QUIC-TLS].
The Initial packet (shown in Figure 5) has two additional header QUIC allows for an arbitrary number of streams to operate
fields that are added to the Long Header before the Length field. concurrently. An endpoint limits the number of concurrently active
incoming streams by limiting the maximum stream ID (see Section 4.5).
+-+-+-+-+-+-+-+-+ The maximum stream ID is specific to each endpoint and applies only
|1| 0x7f | to the peer that receives the setting. That is, clients specify the
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ maximum stream ID the server can initiate, and servers specify the
| Version (32) | maximum stream ID the client can initiate. Each endpoint may respond
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ on streams initiated by the other peer, regardless of whether it is
|DCIL(4)|SCIL(4)| permitted to initiate new streams.
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Source Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Token Length (i) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Token (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Length (i) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Packet Number (8/16/32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Payload (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 5: Initial Packet Endpoints MUST NOT exceed the limit set by their peer. An endpoint
that receives a STREAM frame with an ID greater than the limit it has
sent MUST treat this as a stream error of type STREAM_ID_ERROR
(Section 11), unless this is a result of a change in the initial
limits (see Section 7.3.1).
These fields include the token that was previously provided in a A receiver cannot renege on an advertisement; that is, once a
Retry packet or NEW_TOKEN frame: receiver advertises a stream ID via a MAX_STREAM_ID frame,
advertising a smaller maximum ID has no effect. A receiver MUST
ignore any MAX_STREAM_ID frame that does not increase the maximum
stream ID.
Token Length: A variable-length integer specifying the length of the 2.3. Sending and Receiving Data
Token field, in bytes. This value is zero if no token is present.
Initial packets sent by the server MUST set the Token Length field
to zero; clients that receive an Initial packet with a non-zero
Token Length field MUST either discard the packet or generate a
connection error of type PROTOCOL_VIOLATION.
Token: The value of the token. Endpoints uses streams to send and receive data. Endpoints send
STREAM frames, which encapsulate data for a stream. STREAM frames
carry a flag that can be used to signal the end of a stream.
The client and server use the Initial packet type for any packet that Streams are an ordered byte-stream abstraction with no other
contains an initial cryptographic handshake message. This includes structure that is visible to QUIC. STREAM frame boundaries are not
all cases where a new packet containing the initial cryptographic expected to preserved when data is transmitted, when data is
message needs to be created, such as the packets sent after receiving retransmitted after packet loss, or when data is delivered to the
a Version Negotiation (Section 4.3) or Retry packet (Section 4.4). application at the receiver.
A server sends its first Initial packet in response to a client When new data is to be sent on a stream, a sender MUST set the
Initial. A server may send multiple Initial packets. The encapsulating STREAM frame's offset field to the stream offset of the
cryptographic key exchange could require multiple round trips or first octet of this new data. The first octet of data on a stream
retransmissions of this data. has an offset of 0. An endpoint is expected to send every stream
octet. The largest offset delivered on a stream MUST be less than
2^62.
The payload of an Initial packet includes a CRYPTO frame (or frames) QUIC makes no specific allowances for partial reliability or delivery
containing a cryptographic handshake message, ACK frames, or both. of stream data out of order. Endpoints MUST be able to deliver
PADDING and CONNECTION_CLOSE frames are also permitted. An endpoint stream data to an application as an ordered byte-stream. Delivering
that receives an Initial packet containing other frames can either an ordered byte-stream requires that an endpoint buffer any data that
discard the packet as spurious or treat it as a connection error. is received out of order, up to the advertised flow control limit.
The first packet sent by a client always includes a CRYPTO frame that An endpoint could receive the same octets multiple times; octets that
contains the entirety of the first cryptographic handshake message. have already been received can be discarded. The value for a given
This packet, and the cryptographic handshake message, MUST fit in a octet MUST NOT change if it is sent multiple times; an endpoint MAY
single UDP datagram (see Section 6.4). The first CRYPTO frame sent treat receipt of a changed octet as a connection error of type
always begins at an offset of 0 (see Section 6.4). PROTOCOL_VIOLATION.
Note that if the server sends a HelloRetryRequest, the client will An endpoint MUST NOT send data on any stream without ensuring that it
send a second Initial packet. This Initial packet will continue the is within the data limits set by its peer. Flow control is described
cryptographic handshake and will contain a CRYPTO frame with an in detail in Section 4.
offset matching the size of the CRYPTO frame sent in the first
Initial packet. Cryptographic handshake messages subsequent to the
first do not need to fit within a single UDP datagram.
4.6.1. Connection IDs 2.4. Stream Prioritization
When an Initial packet is sent by a client which has not previously Stream multiplexing has a significant effect on application
received a Retry packet from the server, it populates the Destination performance if resources allocated to streams are correctly
Connection ID field with an unpredictable value. This MUST be at prioritized. Experience with other multiplexed protocols, such as
least 8 octets in length. Until a packet is received from the HTTP/2 [HTTP2], shows that effective prioritization strategies have a
server, the client MUST use the same value unless it abandons the significant positive impact on performance.
connection attempt and starts a new one. The initial Destination
Connection ID is used to determine packet protection keys for Initial
packets.
The client populates the Source Connection ID field with a value of QUIC does not provide frames for exchanging prioritization
its choosing and sets the SCIL field to match. information. Instead it relies on receiving priority information
from the application that uses QUIC. Protocols that use QUIC are
able to define any prioritization scheme that suits their application
semantics. A protocol might define explicit messages for signaling
priority, such as those defined in HTTP/2; it could define rules that
allow an endpoint to determine priority based on context; or it could
leave the determination to the application.
The Destination Connection ID field in the server's Initial packet A QUIC implementation SHOULD provide ways in which an application can
contains a connection ID that is chosen by the recipient of the indicate the relative priority of streams. When deciding which
packet (i.e., the client); the Source Connection ID includes the streams to dedicate resources to, QUIC SHOULD use the information
connection ID that the sender of the packet wishes to use (see provided by the application. Failure to account for priority of
Section 6.1). The server MUST use consistent Source Connection IDs streams can result in suboptimal performance.
during the handshake.
On first receiving an Initial or Retry packet from the server, the Stream priority is most relevant when deciding which stream data will
client uses the Source Connection ID supplied by the server as the be transmitted. Often, there will be limits on what can be
Destination Connection ID for subsequent packets. That means that a transmitted as a result of connection flow control or the current
client might change the Destination Connection ID twice during congestion controller state.
connection establishment. Once a client has received an Initial
packet from the server, it MUST discard any packet it receives with a
different Source Connection ID.
4.6.2. Tokens Giving preference to the transmission of its own management frames
ensures that the protocol functions efficiently. That is,
prioritizing frames other than STREAM frames ensures that loss
recovery, congestion control, and flow control operate effectively.
If the client has a token received in a NEW_TOKEN frame on a previous CRYPTO frames SHOULD be prioritized over other streams prior to the
connection to what it believes to be the same server, it can include completion of the cryptographic handshake. This includes the
that value in the Token field of its Initial packet. retransmission of the second flight of client handshake messages,
that is, the TLS Finished and any client authentication messages.
A token allows a server to correlate activity between connections. STREAM data in frames determined to be lost SHOULD be retransmitted
Specifically, the connection where the token was issued, and any before sending new data, unless application priorities indicate
connection where it is used. Clients that want to break continuity otherwise. Retransmitting lost stream data can fill in gaps, which
of identity with a server MAY discard tokens provided using the allows the peer to consume already received data and free up the flow
NEW_TOKEN frame. Tokens obtained in Retry packets MUST NOT be control window.
discarded.
A client SHOULD NOT reuse a token. Reusing a token allows 3. Stream States: Life of a Stream
connections to be linked by entities on the network path (see
Section 6.11.5). A client MUST NOT reuse a token if it believes that
its point of network attachment has changed since the token was last
used; that is, if there is a change in its local IP address or
network interface. A client needs to start the connection process
over if it migrates prior to completing the handshake.
When a server receives an Initial packet with an address validation This section describes the two types of QUIC stream in terms of the
token, it SHOULD attempt to validate it. If the token is invalid states of their send or receive components. Two state machines are
then the server SHOULD proceed as if the client did not have a described: one for streams on which an endpoint transmits data
validated address, including potentially sending a Retry. If the (Section 3.1); another for streams from which an endpoint receives
validation succeeds, the server SHOULD then allow the handshake to data (Section 3.2).
proceed (see Section 6.7).
Note: The rationale for treating the client as unvalidated rather Unidirectional streams use the applicable state machine directly.
than discarding the packet is that the client might have received Bidirectional streams use both state machines. For the most part,
the token in a previous connection using the NEW_TOKEN frame, and the use of these state machines is the same whether the stream is
if the server has lost state, it might be unable to validate the unidirectional or bidirectional. The conditions for opening a stream
token at all, leading to connection failure if the packet is are slightly more complex for a bidirectional stream because the
discarded. A server MAY encode tokens provided with NEW_TOKEN opening of either send or receive sides causes the stream to open in
frames and Retry packets differently, and validate the latter more both directions.
strictly.
In a stateless design, a server can use encrypted and authenticated An endpoint can open streams up to its maximum stream limit in any
tokens to pass information to clients that the server can later order, however endpoints SHOULD open the send side of streams for
recover and use to validate a client address. Tokens are not each type in order.
integrated into the cryptographic handshake and so they are not
authenticated. For instance, a client might be able to reuse a
token. To avoid attacks that exploit this property, a server can
limit its use of tokens to only the information needed validate
client addresses.
4.6.3. Starting Packet Numbers Note: These states are largely informative. This document uses
stream states to describe rules for when and how different types
of frames can be sent and the reactions that are expected when
different types of frames are received. Though these state
machines are intended to be useful in implementing QUIC, these
states aren't intended to constrain implementations. An
implementation can define a different state machine as long as its
behavior is consistent with an implementation that implements
these states.
The first Initial packet sent by either endpoint contains a packet 3.1. Send Stream States
number of 0. The packet number MUST increase monotonically
thereafter. Initial packets are in a different packet number space
to other packets (see Section 4.11).
4.6.4. 0-RTT Packet Numbers Figure 1 shows the states for the part of a stream that sends data to
a peer.
Packet numbers for 0-RTT protected packets use the same space as o
1-RTT protected packets. | Create Stream (Sending)
| Create Bidirectional Stream (Receiving)
v
+-------+
| Ready | Send RST_STREAM
| |-----------------------.
+-------+ |
| |
| Send STREAM / |
| STREAM_BLOCKED |
| |
| Create Bidirectional |
| Stream (Receiving) |
v |
+-------+ |
| Send | Send RST_STREAM |
| |---------------------->|
+-------+ |
| |
| Send STREAM + FIN |
v v
+-------+ +-------+
| Data | Send RST_STREAM | Reset |
| Sent |------------------>| Sent |
+-------+ +-------+
| |
| Recv All ACKs | Recv ACK
v v
+-------+ +-------+
| Data | | Reset |
| Recvd | | Recvd |
+-------+ +-------+
After a client receives a Retry or Version Negotiation packet, 0-RTT Figure 1: States for Send Streams
packets are likely to have been lost or discarded by the server. A
client MAY attempt to resend data in 0-RTT packets after it sends a
new Initial packet.
A client MUST NOT reset the packet number it uses for 0-RTT packets. The sending part of stream that the endpoint initiates (types 0 and 2
The keys used to protect 0-RTT packets will not change as a result of for clients, 1 and 3 for servers) is opened by the application or
responding to a Retry or Version Negotiation packet unless the client application protocol. The "Ready" state represents a newly created
also regenerates the cryptographic handshake message. Sending stream that is able to accept data from the application. Stream data
packets with the same packet number in that case is likely to might be buffered in this state in preparation for sending.
compromise the packet protection for all 0-RTT packets because the
same key and nonce could be used to protect different content.
Receiving a Retry or Version Negotiation packet, especially a Retry Sending the first STREAM or STREAM_BLOCKED frame causes a send stream
that changes the connection ID used for subsequent packets, indicates to enter the "Send" state. An implementation might choose to defer
a strong possibility that 0-RTT packets could be lost. A client only allocating a Stream ID to a send stream until it sends the first
receives acknowledgments for its 0-RTT packets once the handshake is frame and enters this state, which can allow for better stream
complete. Consequently, a server might expect 0-RTT packets to start prioritization.
with a packet number of 0. Therefore, in determining the length of
the packet number encoding for 0-RTT packets, a client MUST assume
that all packets up to the current packet number are in flight,
starting from a packet number of 0. Thus, 0-RTT packets could need
to use a longer packet number encoding.
A client SHOULD instead generate a fresh cryptographic handshake The sending part of a bidirectional stream initiated by a peer (type
message and start packet numbers from 0. This ensures that new 0-RTT 0 for a server, type 1 for a client) enters the "Ready" state then
packets will not use the same keys, avoiding any risk of key and immediately transitions to the "Send" state if the receiving part
nonce reuse; this also prevents 0-RTT packets from previous handshake enters the "Recv" state.
attempts from being accepted as part of the connection.
4.6.5. Minimum Packet Size In the "Send" state, an endpoint transmits - and retransmits as
necessary - data in STREAM frames. The endpoint respects the flow
control limits of its peer, accepting MAX_STREAM_DATA frames. An
endpoint in the "Send" state generates STREAM_BLOCKED frames if it
encounters flow control limits.
The payload of a UDP datagram carrying the Initial packet MUST be After the application indicates that stream data is complete and a
expanded to at least 1200 octets (see Section 8), by adding PADDING STREAM frame containing the FIN bit is sent, the send stream enters
frames to the Initial packet and/or by combining the Initial packet the "Data Sent" state. From this state, the endpoint only
with a 0-RTT packet (see Section 4.9). retransmits stream data as necessary. The endpoint no longer needs
to track flow control limits or send STREAM_BLOCKED frames for a send
stream in this state. The endpoint can ignore any MAX_STREAM_DATA
frames it receives from its peer in this state; MAX_STREAM_DATA
frames might be received until the peer receives the final stream
offset.
4.7. Handshake Packet Once all stream data has been successfully acknowledged, the send
stream enters the "Data Recvd" state, which is a terminal state.
A Handshake packet uses long headers with a type value of 0x7D. It From any of the "Ready", "Send", or "Data Sent" states, an
is used to carry acknowledgments and cryptographic handshake messages application can signal that it wishes to abandon transmission of
from the server and client. stream data. Similarly, the endpoint might receive a STOP_SENDING
frame from its peer. In either case, the endpoint sends a RST_STREAM
frame, which causes the stream to enter the "Reset Sent" state.
A server sends its cryptographic handshake in one or more Handshake An endpoint MAY send a RST_STREAM as the first frame on a send
packets in response to an Initial packet if it does not send a Retry stream; this causes the send stream to open and then immediately
packet. Once a client has received a Handshake packet from a server, transition to the "Reset Sent" state.
it uses Handshake packets to send subsequent cryptographic handshake
messages and acknowledgments to the server.
The Destination Connection ID field in a Handshake packet contains a Once a packet containing a RST_STREAM has been acknowledged, the send
connection ID that is chosen by the recipient of the packet; the stream enters the "Reset Recvd" state, which is a terminal state.
Source Connection ID includes the connection ID that the sender of
the packet wishes to use (see Section 4.10).
The first Handshake packet sent by a server contains a packet number 3.2. Receive Stream States
of 0. Handshake packets are their own packet number space. Packet
numbers are incremented normally for other Handshake packets.
Servers MUST NOT send more than three times as many bytes as the Figure 2 shows the states for the part of a stream that receives data
number of bytes received prior to verifying the client's address. from a peer. The states for a receive stream mirror only some of the
Source addresses can be verified through an address validation token states of the send stream at the peer. A receive stream doesn't
(delivered via a Retry packet or a NEW_TOKEN frame) or by processing track states on the send stream that cannot be observed, such as the
any message from the client encrypted using the Handshake keys. This "Ready" state; instead, receive streams track the delivery of data to
limit exists to mitigate amplification attacks. the application or application protocol some of which cannot be
observed by the sender.
In order to prevent this limit causing a handshake deadlock, the o
client SHOULD always send a packet upon a handshake timeout, as | Recv STREAM / STREAM_BLOCKED / RST_STREAM
described in [QUIC-RECOVERY]. If the client has no data to | Create Bidirectional Stream (Sending)
retransmit and does not have Handshake keys, it SHOULD send an | Recv MAX_STREAM_DATA
Initial packet in a UDP datagram of at least 1200 octets. If the | Create Higher-Numbered Stream
client has Handshake keys, it SHOULD send a Handshake packet. v
+-------+
| Recv | Recv RST_STREAM
| |-----------------------.
+-------+ |
| |
| Recv STREAM + FIN |
v |
+-------+ |
| Size | Recv RST_STREAM |
| Known |---------------------->|
+-------+ |
| |
| Recv All Data |
v v
+-------+ Recv RST_STREAM +-------+
| Data |--- (optional) --->| Reset |
| Recvd | Recv All Data | Recvd |
+-------+<-- (optional) ----+-------+
| |
| App Read All Data | App Read RST
v v
+-------+ +-------+
| Data | | Reset |
| Read | | Read |
+-------+ +-------+
The payload of this packet contains CRYPTO frames and could contain Figure 2: States for Receive Streams
PADDING, or ACK frames. Handshake packets MAY contain
CONNECTION_CLOSE or APPLICATION_CLOSE frames. Endpoints MUST treat
receipt of Handshake packets with other frames as a connection error.
4.8. Protected Packets The receiving part of a stream initiated by a peer (types 1 and 3 for
a client, or 0 and 2 for a server) are created when the first STREAM,
STREAM_BLOCKED, RST_STREAM, or MAX_STREAM_DATA (bidirectional only,
see below) is received for that stream. The initial state for a
receive stream is "Recv". Receiving a RST_STREAM frame causes the
receive stream to immediately transition to the "Reset Recvd".
All QUIC packets use packet protection. Packets that are protected The receive stream enters the "Recv" state when the sending part of a
with the static handshake keys or the 0-RTT keys are sent with long bidirectional stream initiated by the endpoint (type 0 for a client,
headers; all packets protected with 1-RTT keys are sent with short type 1 for a server) enters the "Ready" state.
headers. The different packet types explicitly indicate the
encryption level and therefore the keys that are used to remove
packet protection. 0-RTT and 1-RTT protected packets share a single
packet number space.
Packets protected with handshake keys only use packet protection to A bidirectional stream also opens when a MAX_STREAM_DATA frame is
ensure that the sender of the packet is on the network path. This received. Receiving a MAX_STREAM_DATA frame implies that the remote
packet protection is not effective confidentiality protection; any peer has opened the stream and is providing flow control credit. A
entity that receives the Initial packet from a client can recover the MAX_STREAM_DATA frame might arrive before a STREAM or STREAM_BLOCKED
keys necessary to remove packet protection or to generate packets frame if packets are lost or reordered.
that will be successfully authenticated.
Packets protected with 0-RTT and 1-RTT keys are expected to have Before creating a stream, all lower-numbered streams of the same type
confidentiality and data origin authentication; the cryptographic MUST be created. That means that receipt of a frame that would open
handshake ensures that only the communicating endpoints receive the a stream causes all lower-numbered streams of the same type to be
corresponding keys. opened in numeric order. This ensures that the creation order for
streams is consistent on both endpoints.
Packets protected with 0-RTT keys use a type value of 0x7C. The In the "Recv" state, the endpoint receives STREAM and STREAM_BLOCKED
connection ID fields for a 0-RTT packet MUST match the values used in frames. Incoming data is buffered and can be reassembled into the
the Initial packet (Section 4.6). correct order for delivery to the application. As data is consumed
by the application and buffer space becomes available, the endpoint
sends MAX_STREAM_DATA frames to allow the peer to send more data.
The version field for protected packets is the current QUIC version. When a STREAM frame with a FIN bit is received, the final offset (see
Section 4.3) is known. The receive stream enters the "Size Known"
state. In this state, the endpoint no longer needs to send
MAX_STREAM_DATA frames, it only receives any retransmissions of
stream data.
The packet number field contains a packet number, which has Once all data for the stream has been received, the receive stream
additional confidentiality protection that is applied after packet enters the "Data Recvd" state. This might happen as a result of
protection is applied (see [QUIC-TLS] for details). The underlying receiving the same STREAM frame that causes the transition to "Size
packet number increases with each packet sent, see Section 4.11 for Known". In this state, the endpoint has all stream data. Any STREAM
details. or STREAM_BLOCKED frames it receives for the stream can be discarded.
The payload is protected using authenticated encryption. [QUIC-TLS] The "Data Recvd" state persists until stream data has been delivered
describes packet protection in detail. After decryption, the to the application or application protocol. Once stream data has
plaintext consists of a sequence of frames, as described in been delivered, the stream enters the "Data Read" state, which is a
Section 5. terminal state.
4.9. Coalescing Packets Receiving a RST_STREAM frame in the "Recv" or "Size Known" states
causes the stream to enter the "Reset Recvd" state. This might cause
the delivery of stream data to the application to be interrupted.
A sender can coalesce multiple QUIC packets (typically a It is possible that all stream data is received when a RST_STREAM is
Cryptographic Handshake packet and a Protected packet) into one UDP received (that is, from the "Data Recvd" state). Similarly, it is
datagram. This can reduce the number of UDP datagrams needed to send possible for remaining stream data to arrive after receiving a
application data during the handshake and immediately afterwards. It RST_STREAM frame (the "Reset Recvd" state). An implementation is
is not necessary for senders to coalesce packets, though failing to able to manage this situation as they choose. Sending RST_STREAM
do so will require sending a significantly larger number of datagrams means that an endpoint cannot guarantee delivery of stream data;
during the handshake. Receivers MUST be able to process coalesced however there is no requirement that stream data not be delivered if
packets. a RST_STREAM is received. An implementation MAY interrupt delivery
of stream data, discard any data that was not consumed, and signal
the existence of the RST_STREAM immediately. Alternatively, the
RST_STREAM signal might be suppressed or withheld if stream data is
completely received. In the latter case, the receive stream
effectively transitions to "Data Recvd" from "Reset Recvd".
Coalescing packets in order of increasing encryption levels (Initial, Once the application has been delivered the signal indicating that
0-RTT, Handshake, 1-RTT) makes it more likely the receiver will be the receive stream was reset, the receive stream transitions to the
able to process all the packets in a single pass. A packet with a "Reset Read" state, which is a terminal state.
short header does not include a length, so it will always be the last
packet included in a UDP datagram.
Senders MUST NOT coalesce QUIC packets with different Destination 3.3. Permitted Frame Types
Connection IDs into a single UDP datagram. Receivers SHOULD ignore
any subsequent packets with a different Destination Connection ID
than the first packet in the datagram.
Every QUIC packet that is coalesced into a single UDP datagram is The sender of a stream sends just three frame types that affect the
separate and complete. Though the values of some fields in the state of a stream at either sender or receiver: STREAM
packet header might be redundant, no fields are omitted. The (Section 19.19), STREAM_BLOCKED (Section 19.10), and RST_STREAM
receiver of coalesced QUIC packets MUST individually process each (Section 19.2).
QUIC packet and separately acknowledge them, as if they were received
as the payload of different UDP datagrams. If one or more packets in
a datagram cannot be processed yet (because the keys are not yet
available) or processing fails (decryption failure, unknown type,
etc.), the receiver MUST still attempt to process the remaining
packets. The skipped packets MAY either be discarded or buffered for
later processing, just as if the packets were received out-of-order
in separate datagrams.
Retry (Section 4.4) and Version Negotiation (Section 4.3) packets A sender MUST NOT send any of these frames from a terminal state
cannot be coalesced. ("Data Recvd" or "Reset Recvd"). A sender MUST NOT send STREAM or
STREAM_BLOCKED after sending a RST_STREAM; that is, in the "Reset
Sent" state in addition to the terminal states. A receiver could
receive any of these frames in any state, but only due to the
possibility of delayed delivery of packets carrying them.
4.10. Connection ID Encoding The receiver of a stream sends MAX_STREAM_DATA (Section 19.6) and
STOP_SENDING frames (Section 19.14).
A connection ID is used to ensure consistent routing of packets, as The receiver only sends MAX_STREAM_DATA in the "Recv" state. A
described in Section 6.1. The long header contains two connection receiver can send STOP_SENDING in any state where it has not received
IDs: the Destination Connection ID is chosen by the recipient of the a RST_STREAM frame; that is states other than "Reset Recvd" or "Reset
packet and is used to provide consistent routing; the Source Read". However there is little value in sending a STOP_SENDING frame
Connection ID is used to set the Destination Connection ID used by after all stream data has been received in the "Data Recvd" state. A
the peer. sender could receive these frames in any state as a result of delayed
delivery of packets.
During the handshake, packets with the long header are used to 3.4. Bidirectional Stream States
establish the connection ID that each endpoint uses. Each endpoint
uses the Source Connection ID field to specify the connection ID that
is used in the Destination Connection ID field of packets being sent
to them. Upon receiving a packet, each endpoint sets the Destination
Connection ID it sends to match the value of the Source Connection ID
that they receive.
During the handshake, a client can receive both a Retry and an A bidirectional stream is composed of a send stream and a receive
Initial packet, and thus be given two opportunities to update the stream. Implementations may represent states of the bidirectional
Destination Connection ID it sends. A client MUST only change the stream as composites of send and receive stream states. The simplest
value it sends in the Destination Connection ID in response to the model presents the stream as "open" when either send or receive
first packet of each type it receives from the server (Retry or stream is in a non-terminal state and "closed" when both send and
Initial); a server MUST set its value based on the Initial packet. receive streams are in a terminal state.
Any additional changes are not permitted; if subsequent packets of
those types include a different Source Connection ID, they MUST be
discarded. This avoids problems that might arise from stateless
processing of multiple Initial packets producing different connection
IDs.
Short headers only include the Destination Connection ID and omit the Table 2 shows a more complex mapping of bidirectional stream states
explicit length. The length of the Destination Connection ID field that loosely correspond to the stream states in HTTP/2 [HTTP2]. This
is expected to be known to endpoints. shows that multiple states on send or receive streams are mapped to
the same composite state. Note that this is just one possibility for
such a mapping; this mapping requires that data is acknowledged
before the transition to a "closed" or "half-closed" state.
Endpoints using a connection-ID based load balancer could agree with +-----------------------+---------------------+---------------------+
the load balancer on a fixed or minimum length and on an encoding for | Send Stream | Receive Stream | Composite State |
connection IDs. This fixed portion could encode an explicit length, +-----------------------+---------------------+---------------------+
which allows the entire connection ID to vary in length and still be | No Stream/Ready | No Stream/Recv *1 | idle |
used by the load balancer. | | | |
| Ready/Send/Data Sent | Recv/Size Known | open |
| | | |
| Ready/Send/Data Sent | Data Recvd/Data | half-closed |
| | Read | (remote) |
| | | |
| Ready/Send/Data Sent | Reset Recvd/Reset | half-closed |
| | Read | (remote) |
| | | |
| Data Recvd | Recv/Size Known | half-closed (local) |
| | | |
| Reset Sent/Reset | Recv/Size Known | half-closed (local) |
| Recvd | | |
| | | |
| Data Recvd | Recv/Size Known | half-closed (local) |
| | | |
| Reset Sent/Reset | Data Recvd/Data | closed |
| Recvd | Read | |
| | | |
| Reset Sent/Reset | Reset Recvd/Reset | closed |
| Recvd | Read | |
| | | |
| Data Recvd | Data Recvd/Data | closed |
| | Read | |
| | | |
| Data Recvd | Reset Recvd/Reset | closed |
| | Read | |
+-----------------------+---------------------+---------------------+
The very first packet sent by a client includes a random value for Table 2: Possible Mapping of Stream States to HTTP/2
Destination Connection ID. The same value MUST be used for all 0-RTT
packets sent on that connection (Section 4.8). This randomized value
is used to determine the packet protection keys for Initial packets
(see Section 5.2 of [QUIC-TLS]).
A Version Negotiation (Section 4.3) packet MUST use both connection Note (*1): A stream is considered "idle" if it has not yet been
IDs selected by the client, swapped to ensure correct routing toward created, or if the receive stream is in the "Recv" state without
the client. yet having received any frames.
The connection ID can change over the lifetime of a connection, 3.5. Solicited State Transitions
especially in response to connection migration (Section 6.11).
NEW_CONNECTION_ID frames (Section 7.13) are used to provide new
connection ID values.
4.11. Packet Numbers If an endpoint is no longer interested in the data it is receiving on
a stream, it MAY send a STOP_SENDING frame identifying that stream to
prompt closure of the stream in the opposite direction. This
typically indicates that the receiving application is no longer
reading data it receives from the stream, but is not a guarantee that
incoming data will be ignored.
The packet number is an integer in the range 0 to 2^62-1. The value STREAM frames received after sending STOP_SENDING are still counted
is used in determining the cryptographic nonce for packet protection. toward the connection and stream flow-control windows, even though
Each endpoint maintains a separate packet number for sending and these frames will be discarded upon receipt. This avoids potential
receiving. ambiguity about which STREAM frames count toward flow control.
Packet numbers are divided into 3 spaces in QUIC: A STOP_SENDING frame requests that the receiving endpoint send a
RST_STREAM frame. An endpoint that receives a STOP_SENDING frame
MUST send a RST_STREAM frame for that stream, and can use an error
code of STOPPING. If the STOP_SENDING frame is received on a send
stream that is already in the "Data Sent" state, a RST_STREAM frame
MAY still be sent in order to cancel retransmission of previously-
sent STREAM frames.
o Initial space: All Initial packets Section 4.6 are in this space. STOP_SENDING SHOULD only be sent for a receive stream that has not
been reset. STOP_SENDING is most useful for streams in the "Recv" or
"Size Known" states.
o Handshake space: All Handshake packets Section 4.7 are in this An endpoint is expected to send another STOP_SENDING frame if a
space. packet containing a previous STOP_SENDING is lost. However, once
either all stream data or a RST_STREAM frame has been received for
the stream - that is, the stream is in any state other than "Recv" or
"Size Known" - sending a STOP_SENDING frame is unnecessary.
o Application data space: All 0-RTT and 1-RTT encrypted packets 4. Flow Control
Section 4.8 are in this space.
As described in [QUIC-TLS], each packet type uses different It is necessary to limit the amount of data that a sender may have
protection keys. outstanding at any time, so as to prevent a fast sender from
overwhelming a slow receiver, or to prevent a malicious sender from
consuming significant resources at a receiver. To this end, QUIC
employs a credit-based flow-control scheme similar to that in HTTP/2
[HTTP2]. A receiver advertises the number of octets it is prepared
to receive on a given stream and for the entire connection. This
leads to two levels of flow control in QUIC:
Conceptually, a packet number space is the context in which a packet o Stream flow control, which prevents a single stream from consuming
can be processed and acknowledged. Initial packets can only be sent the entire receive buffer for a connection.
with Initial packet protection keys and acknowledged in packets which
are also Initial packets. Similarly, Handshake packets are sent at
the Handshake encryption level and can only be acknowledged in
Handshake packets.
This enforces cryptographic separation between the data sent in the o Connection flow control, which prevents senders from exceeding a
different packet sequence number spaces. Each packet number space receiver's buffer capacity for the connection, and
starts at packet number 0. Subsequent packets sent in the same
packet number space MUST increase the packet number by at least one.
0-RTT and 1-RTT data exist in the same packet number space to make A data receiver sets initial credits for all streams by sending
loss recovery algorithms easier to implement between the two packet transport parameters during the handshake (Section 7.3).
types.
A QUIC endpoint MUST NOT reuse a packet number within the same packet A data receiver sends MAX_STREAM_DATA or MAX_DATA frames to the
number space in one connection (that is, under the same cryptographic sender to advertise additional credit. MAX_STREAM_DATA frames send
keys). If the packet number for sending reaches 2^62 - 1, the sender the maximum absolute byte offset of a stream, while MAX_DATA frames
MUST close the connection without sending a CONNECTION_CLOSE frame or send the maximum of the sum of the absolute byte offsets of all
any further packets; an endpoint MAY send a Stateless Reset streams.
(Section 6.13.4) in response to further packets that it receives.
In the QUIC long and short packet headers, the number of bits A receiver advertises credit for a stream by sending a
required to represent the packet number is reduced by including only MAX_STREAM_DATA frame with the Stream ID set appropriately. A
a variable number of the least significant bits of the packet number. receiver could use the current offset of data consumed to determine
One or two of the most significant bits of the first octet determine the flow control offset to be advertised. A receiver MAY send
how many bits of the packet number are provided, as shown in Table 2. MAX_STREAM_DATA frames in multiple packets in order to make sure that
the sender receives an update before running out of flow control
credit, even if one of the packets is lost.
+---------------------+----------------+--------------+ Connection flow control is a limit to the total bytes of stream data
| First octet pattern | Encoded Length | Bits Present | sent in STREAM frames on all streams. A receiver advertises credit
+---------------------+----------------+--------------+ for a connection by sending a MAX_DATA frame. A receiver maintains a
| 0b0xxxxxxx | 1 octet | 7 | cumulative sum of bytes received on all contributing streams, which
| | | | are used to check for flow control violations. A receiver might use
| 0b10xxxxxx | 2 | 14 | a sum of bytes consumed on all streams to determine the maximum data
| | | | limit to be advertised.
| 0b11xxxxxx | 4 | 30 |
+---------------------+----------------+--------------+
Table 2: Packet Number Encodings for Packet Headers A receiver MAY advertise a larger offset at any point by sending
MAX_STREAM_DATA or MAX_DATA frames. A receiver cannot renege on an
advertisement; that is, once a receiver advertises an offset,
advertising a smaller offset has no effect. A sender MUST therefore
ignore any MAX_STREAM_DATA or MAX_DATA frames that do not increase
flow control limits.
Note that these encodings are similar to those in Section 7.1, but A receiver MUST close the connection with a FLOW_CONTROL_ERROR error
use different values. (Section 11) if the peer violates the advertised connection or stream
data limits.
The encoded packet number is protected as described in Section 5.3 A sender SHOULD send STREAM_BLOCKED or BLOCKED frames to indicate it
[QUIC-TLS]. Protection of the packet number is removed prior to has data to write but is blocked by flow control limits. These
recovering the full packet number. The full packet number is frames are expected to be sent infrequently in common cases, but they
reconstructed at the receiver based on the number of significant bits are considered useful for debugging and monitoring purposes.
present, the value of those bits, and the largest packet number
received on a successfully authenticated packet. Recovering the full
packet number is necessary to successfully remove packet protection.
Once packet number protection is removed, the packet number is A similar method is used to control the number of open streams (see
decoded by finding the packet number value that is closest to the Section 4.5 for details).
next expected packet. The next expected packet is the highest
received packet number plus one. For example, if the highest
successfully authenticated packet had a packet number of 0xaa82f30e,
then a packet containing a 14-bit value of 0x9b3 will be decoded as
0xaa8309b3. Example pseudo-code for packet number decoding can be
found in Appendix A.
The sender MUST use a packet number size able to represent more than 4.1. Handling of Stream Cancellation
twice as large a range than the difference between the largest
acknowledged packet and packet number being sent. A peer receiving
the packet will then correctly decode the packet number, unless the
packet is delayed in transit such that it arrives after many higher-
numbered packets have been received. An endpoint SHOULD use a large
enough packet number encoding to allow the packet number to be
recovered even if the packet arrives after packets that are sent
afterwards.
As a result, the size of the packet number encoding is at least one There are some edge cases which must be considered when dealing with
more than the base 2 logarithm of the number of contiguous stream and connection level flow control. Given enough time, both
unacknowledged packet numbers, including the new packet. endpoints must agree on flow control state. If one end believes it
can send more than the other end is willing to receive, the
connection will be torn down when too much data arrives. Conversely
if a sender believes it is blocked, while endpoint B expects more
data can be received, then the connection can be in a deadlock, with
the sender waiting for a MAX_STREAM_DATA or MAX_DATA frame which will
never come.
For example, if an endpoint has received an acknowledgment for packet On receipt of a RST_STREAM frame, an endpoint will tear down state
0x6afa2f, sending a packet with a number of 0x6b2d79 requires a for the matching stream and ignore further data arriving on that
packet number encoding with 14 bits or more; whereas the 30-bit stream. This could result in the endpoints getting out of sync,
packet number encoding is needed to send a packet with a number of since the RST_STREAM frame may have arrived out of order and there
0x6bc107. may be further bytes in flight. The data sender would have counted
the data against its connection level flow control budget, but a
receiver that has not received these bytes would not know to include
them as well. The receiver must learn the number of bytes that were
sent on the stream to make the same adjustment in its connection flow
controller.
A receiver MUST discard a newly unprotected packet unless it is To ensure that endpoints maintain a consistent connection-level flow
certain that it has not processed another packet with the same packet control state, the RST_STREAM frame (Section 19.2) includes the
number from the same packet number space. Duplicate suppression MUST largest offset of data sent on the stream. On receiving a RST_STREAM
happen after removing packet protection for the reasons described in frame, a receiver definitively knows how many bytes were sent on that
Section 9.3 of [QUIC-TLS]. An efficient algorithm for duplicate stream before the RST_STREAM frame, and the receiver MUST use the
suppression can be found in Section 3.4.3 of [RFC2406]. final offset to account for all bytes sent on the stream in its
connection level flow controller.
A Version Negotiation packet (Section 4.3) does not include a packet RST_STREAM terminates one direction of a stream abruptly. Whether
number. The Retry packet (Section 4.4) has special rules for any action or response can or should be taken on the data already
populating the packet number field. received is application specific.
5. Frames and Frame Types For a bidirectional stream, RST_STREAM has no effect on data flow in
the opposite direction. The RST_STREAM sender can send a
STOP_SENDING frame to encourage prompt termination. Both endpoints
MUST maintain state for the stream in the unterminated direction
until that direction enters a terminal state, or either side sends
CONNECTION_CLOSE or APPLICATION_CLOSE.
The payload of all packets, after removing packet protection, 4.2. Data Limit Increments
consists of a sequence of frames, as shown in Figure 6. Version
Negotiation and Stateless Reset do not contain frames.
0 1 2 3 This document leaves when and how many bytes to advertise in a
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 MAX_DATA or MAX_STREAM_DATA to implementations, but offers a few
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ considerations. These frames contribute to connection overhead.
| Frame 1 (*) ... Therefore frequently sending frames with small changes is
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ undesirable. At the same time, larger increments to limits are
| Frame 2 (*) ... necessary to avoid blocking if updates are less frequent, requiring
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ larger resource commitments at the receiver. Thus there is a trade-
... off between resource commitment and overhead when determining how
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ large a limit is advertised.
| Frame N (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 6: QUIC Payload A receiver MAY use an autotuning mechanism to tune the frequency and
amount that it increases data limits based on a round-trip time
estimate and the rate at which the receiving application consumes
data, similar to common TCP implementations.
QUIC payloads MUST contain at least one frame, and MAY contain If a sender runs out of flow control credit, it will be unable to
multiple frames and multiple frame types. send new data. That is, the sender is blocked. A blocked sender
SHOULD send a STREAM_BLOCKED or BLOCKED frame. A receiver uses these
frames for debugging purposes. A receiver MUST NOT wait for a
STREAM_BLOCKED or BLOCKED frame before sending MAX_STREAM_DATA or
MAX_DATA, since doing so will mean that a sender will be blocked for
an entire round trip and the peer may never send a STREAM_BLOCKED or
BLOCKED frame.
Frames MUST fit within a single QUIC packet and MUST NOT span a QUIC It is generally considered best to not let the sender go into
packet boundary. Each frame begins with a Frame Type, indicating its quiescence if avoidable. To avoid blocking a sender, and to
type, followed by additional type-dependent fields: reasonably account for the possibility of loss, a receiver should
send a MAX_DATA or MAX_STREAM_DATA frame at least two round trips
before it expects the sender to get blocked.
0 1 2 3 A sender sends a single BLOCKED or STREAM_BLOCKED frame only once
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 when it reaches a data limit. A sender SHOULD NOT send multiple
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ BLOCKED or STREAM_BLOCKED frames for the same data limit, unless the
| Frame Type (i) ... original frame is determined to be lost. Another BLOCKED or
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ STREAM_BLOCKED frame can be sent after the data limit is increased.
| Type-Dependent Fields (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 7: Generic Frame Layout 4.3. Stream Final Offset
The frame types defined in this specification are listed in Table 3. The final offset is the count of the number of octets that are
The Frame Type in STREAM frames is used to carry other frame-specific transmitted on a stream. For a stream that is reset, the final
flags. For all other frames, the Frame Type field simply identifies offset is carried explicitly in a RST_STREAM frame. Otherwise, the
the frame. These frames are explained in more detail as they are final offset is the offset of the end of the data carried in a STREAM
referenced later in the document. frame marked with a FIN flag, or 0 in the case of incoming
unidirectional streams.
+-------------+----------------------+--------------+ An endpoint will know the final offset for a stream when the receive
| Type Value | Frame Type Name | Definition | stream enters the "Size Known" or "Reset Recvd" state.
+-------------+----------------------+--------------+
| 0x00 | PADDING | Section 7.2 |
| | | |
| 0x01 | RST_STREAM | Section 7.3 |
| | | |
| 0x02 | CONNECTION_CLOSE | Section 7.4 |
| | | |
| 0x03 | APPLICATION_CLOSE | Section 7.5 |
| | | |
| 0x04 | MAX_DATA | Section 7.6 |
| | | |
| 0x05 | MAX_STREAM_DATA | Section 7.7 |
| | | |
| 0x06 | MAX_STREAM_ID | Section 7.8 |
| | | |
| 0x07 | PING | Section 7.9 |
| | | |
| 0x08 | BLOCKED | Section 7.10 |
| | | |
| 0x09 | STREAM_BLOCKED | Section 7.11 |
| | | |
| 0x0a | STREAM_ID_BLOCKED | Section 7.12 |
| | | |
| 0x0b | NEW_CONNECTION_ID | Section 7.13 |
| | | |
| 0x0c | STOP_SENDING | Section 7.15 |
| | | |
| 0x0d | RETIRE_CONNECTION_ID | Section 7.14 |
| | | |
| 0x0e | PATH_CHALLENGE | Section 7.17 |
| | | |
| 0x0f | PATH_RESPONSE | Section 7.18 |
| | | |
| 0x10 - 0x17 | STREAM | Section 7.20 |
| | | |
| 0x18 | CRYPTO | Section 7.21 |
| | | |
| 0x19 | NEW_TOKEN | Section 7.19 |
| | | |
| 0x1a - 0x1b | ACK | Section 7.16 |
+-------------+----------------------+--------------+
Table 3: Frame Types An endpoint MUST NOT send data on a stream at or beyond the final
offset.
All QUIC frames are idempotent. That is, a valid frame does not Once a final offset for a stream is known, it cannot change. If a
cause undesirable side effects or errors when received more than RST_STREAM or STREAM frame causes the final offset to change for a
once. stream, an endpoint SHOULD respond with a FINAL_OFFSET_ERROR error
(see Section 11). A receiver SHOULD treat receipt of data at or
beyond the final offset as a FINAL_OFFSET_ERROR error, even after a
stream is closed. Generating these errors is not mandatory, but only
because requiring that an endpoint generate these errors also means
that the endpoint needs to maintain the final offset state for closed
streams, which could mean a significant state commitment.
The Frame Type field uses a variable length integer encoding (see 4.4. Flow Control for Cryptographic Handshake
Section 7.1) with one exception. To ensure simple and efficient
implementations of frame parsing, a frame type MUST use the shortest
possible encoding. Though a two-, four- or eight-octet encoding of
the frame types defined in this document is possible, the Frame Type
field for these frames is encoded on a single octet. For instance,
though 0x4007 is a legitimate two-octet encoding for a variable-
length integer with a value of 7, PING frames are always encoded as a
single octet with the value 0x07. An endpoint MUST treat the receipt
of a frame type that uses a longer encoding than necessary as a
connection error of type PROTOCOL_VIOLATION.
5.1. Extension Frames Data sent in CRYPTO frames is not flow controlled in the same way as
STREAM frames. QUIC relies on the cryptographic protocol
implementation to avoid excessive buffering of data, see [QUIC-TLS].
The implementation SHOULD provide an interface to QUIC to tell it
about its buffering limits so that there is not excessive buffering
at multiple layers.
QUIC frames do not use a self-describing encoding. An endpoint 4.5. Stream Limit Increment
therefore needs to understand the syntax of all frames before it can
successfully process a packet. This allows for efficient encoding of
frames, but it means that an endpoint cannot send a frame of a type
that is unknown to its peer.
An extension to QUIC that wishes to use a new type of frame MUST An endpoint limits the number of concurrently active incoming streams
first ensure that a peer is able to understand the frame. An by limiting the maximum stream ID. An initial value is set in the
endpoint can use a transport parameter to signal its willingness to transport parameters (see Section 18.1) and is subsequently increased
receive one or more extension frame types with the one transport by MAX_STREAM_ID frames (see Section 19.7).
parameter.
Extension frames MUST be congestion controlled and MUST cause an ACK As with stream and connection flow control, this document leaves when
frame to be sent. The exception is extension frames that replace or and how many streams to make available to a peer via MAX_STREAM_ID to
supplement the ACK frame. Extension frames are not included in flow implementations, but offers a few considerations. MAX_STREAM_ID
control unless specified in the extension. frames constitute minimal overhead, while withholding MAX_STREAM_ID
frames can prevent the peer from using the available parallelism.
An IANA registry is used to manage the assignment of frame types, see The STREAM_ID_BLOCKED frame (Section 19.11) can be used to signal a
Section 13.2. shortage of available streams. Implementations will likely want to
increase the maximum stream ID as peer-initiated streams close.
6. Life of a Connection 5. Connections
A QUIC connection is a single conversation between two QUIC A QUIC connection is a single conversation between two QUIC
endpoints. QUIC's connection establishment intertwines version endpoints. QUIC's connection establishment combines version
negotiation with the cryptographic and transport handshakes to reduce negotiation with the cryptographic and transport handshakes to reduce
connection establishment latency, as described in Section 6.4. Once connection establishment latency, as described in Section 7. Once
established, a connection may migrate to a different IP or port at established, a connection may migrate to a different IP or port at
either endpoint, due to NAT rebinding or mobility, as described in either endpoint as described in Section 9. Finally, a connection may
Section 6.11. Finally, a connection may be terminated by either be terminated by either endpoint, as described in Section 10.
endpoint, as described in Section 6.13.
6.1. Connection ID 5.1. Connection ID
Each connection possesses a set of identifiers, any of which could be Each connection possesses a set of connection identifiers, or
used to distinguish it from other connections. Connection IDs are connection IDs, each of which can be identify the connection.
selected independently in each direction. Each Connection ID has an Connection IDs are independently selected by endpoints; each endpoint
associated sequence number to assist in deduplicating messages. selects the connection IDs that its peer uses.
The primary function of a connection ID is to ensure that changes in The primary function of a connection ID is to ensure that changes in
addressing at lower protocol layers (UDP, IP, and below) don't cause addressing at lower protocol layers (UDP, IP, and below) don't cause
packets for a QUIC connection to be delivered to the wrong endpoint. packets for a QUIC connection to be delivered to the wrong endpoint.
Each endpoint selects connection IDs using an implementation-specific Each endpoint selects connection IDs using an implementation-specific
(and perhaps deployment-specific) method which will allow packets (and perhaps deployment-specific) method which will allow packets
with that connection ID to be routed back to the endpoint and with that connection ID to be routed back to the endpoint and
identified by the endpoint upon receipt. identified by the endpoint upon receipt.
Connection IDs MUST NOT contain any information that can be used to Connection IDs MUST NOT contain any information that can be used to
correlate them with other connection IDs for the same connection. As correlate them with other connection IDs for the same connection. As
a trivial example, this means the same connection ID MUST NOT be a trivial example, this means the same connection ID MUST NOT be
issued more than once on the same connection. issued more than once on the same connection.
Packets with long headers include Source Connection ID and
Destination Connection ID fields. These fields are used to set the
connection IDs for new connections, see Section 7.2 for details.
Packets with short headers (Section 17.3) only include the
Destination Connection ID and omit the explicit length. The length
of the Destination Connection ID field is expected to be known to
endpoints. Endpoints using a load balancer that routes based on
connection ID could agree with the load balancer on a fixed length
for connection IDs, or agree on an encoding scheme. A fixed portion
could encode an explicit length, which allows the entire connection
ID to vary in length and still be used by the load balancer.
A Version Negotiation (Section 17.4) packet echoes the connection IDs
selected by the client, both to ensure correct routing toward the
client and to allow the client to validate that the packet is in
response to an Initial packet.
A zero-length connection ID MAY be used when the connection ID is not A zero-length connection ID MAY be used when the connection ID is not
needed for routing and the address/port tuple of packets is needed for routing and the address/port tuple of packets is
sufficient to identify a connection. An endpoint whose peer has sufficient to identify a connection. An endpoint whose peer has
selected a zero-length connection ID MUST continue to use a zero- selected a zero-length connection ID MUST continue to use a zero-
length connection ID for the lifetime of the connection and MUST NOT length connection ID for the lifetime of the connection and MUST NOT
send packets from any other local address. send packets from any other local address.
When an endpoint has requested a non-zero-length connection ID, it When an endpoint has requested a non-zero-length connection ID, it
needs to ensure that the peer has a supply of connection IDs from needs to ensure that the peer has a supply of connection IDs from
which to choose for packets sent to the endpoint. These connection which to choose for packets sent to the endpoint. These connection
IDs are supplied by the endpoint using the NEW_CONNECTION_ID frame IDs are supplied by the endpoint using the NEW_CONNECTION_ID frame
(Section 7.13). (Section 19.12).
6.1.1. Issuing Connection IDs 5.1.1. Issuing Connection IDs
The initial connection ID issued by an endpoint is the Source Each Connection ID has an associated sequence number to assist in
Connection ID during the handshake. The sequence number of the deduplicating messages. The initial connection ID issued by an
initial connection ID is 0. If the preferred_address transport endpoint is sent in the Source Connection ID field of the long packet
header (Section 17.2) during the handshake. The sequence number of
the initial connection ID is 0. If the preferred_address transport
parameter is sent, the sequence number of the supplied connection ID parameter is sent, the sequence number of the supplied connection ID
is 1. Subsequent connection IDs are communicated to the peer using is 1.
NEW_CONNECTION_ID frames (Section 7.13), and the sequence number on
Additional connection IDs are communicated to the peer using
NEW_CONNECTION_ID frames (Section 19.12). The sequence number on
each newly-issued connection ID MUST increase by 1. The connection each newly-issued connection ID MUST increase by 1. The connection
ID randomly selected by the client in the Initial packet and any ID randomly selected by the client in the Initial packet and any
connection ID provided by a Reset packet are not assigned sequence connection ID provided by a Reset packet are not assigned sequence
numbers unless a server opts to retain them as its initial connection numbers unless a server opts to retain them as its initial connection
ID. ID.
When an endpoint issues a connection ID, it MUST accept packets that When an endpoint issues a connection ID, it MUST accept packets that
carry this connection ID for the duration of the connection or until carry this connection ID for the duration of the connection or until
its peer invalidates the connection ID via a RETIRE_CONNECTION_ID its peer invalidates the connection ID via a RETIRE_CONNECTION_ID
frame (Section 7.14). frame (Section 19.13).
An endpoint SHOULD ensure that its peer has a sufficient number of An endpoint SHOULD ensure that its peer has a sufficient number of
available and unused connection IDs. While each endpoint available and unused connection IDs. While each endpoint
independently chooses how many connection IDs to issue, endpoints independently chooses how many connection IDs to issue, endpoints
SHOULD provide and maintain at least eight connection IDs. The SHOULD provide and maintain at least eight connection IDs. The
endpoint can do this by always supplying a new connection ID when a endpoint can do this by always supplying a new connection ID when a
connection ID is retired by its peer or when the endpoint receives a connection ID is retired by its peer or when the endpoint receives a
packet with a previously unused connection ID. Endpoints that packet with a previously unused connection ID. Endpoints that
initiate migration and require non-zero-length connection IDs SHOULD initiate migration and require non-zero-length connection IDs SHOULD
provide their peers with new connection IDs before migration, or risk provide their peers with new connection IDs before migration, or risk
the peer closing the connection. the peer closing the connection.
6.1.2. Consuming and Retiring Connection IDs 5.1.2. Consuming and Retiring Connection IDs
An endpoint can change the connection ID it uses for a peer to An endpoint can change the connection ID it uses for a peer to
another available one at any time during the connection. An endpoint another available one at any time during the connection. An endpoint
consumes connection IDs in response to a migrating peer, see consumes connection IDs in response to a migrating peer, see
Section 6.11.5 for more. Section 9.5 for more.
An endpoint maintains a set of connection IDs received from its peer, An endpoint maintains a set of connection IDs received from its peer,
any of which it can use when sending packets. When the endpoint any of which it can use when sending packets. When the endpoint
wishes to remove a connection ID from use, it sends a wishes to remove a connection ID from use, it sends a
RETIRE_CONNECTION_ID frame to its peer, indicating that the peer RETIRE_CONNECTION_ID frame to its peer, indicating that the peer
might bring a new connection ID into circulation using the might bring a new connection ID into circulation using the
NEW_CONNECTION_ID frame. NEW_CONNECTION_ID frame.
An endpoint that retires a connection ID can retain knowledge of that An endpoint that retires a connection ID can retain knowledge of that
connection ID for a period of time after sending the connection ID for a period of time after sending the
RETIRE_CONNECTION_ID frame, or until that frame is acknowledged. RETIRE_CONNECTION_ID frame, or until that frame is acknowledged.
As discussed in Section 6.11.5, each connection ID MUST be used on As discussed in Section 9.5, each connection ID MUST be used on
packets sent from only one local address. An endpoint that migrates packets sent from only one local address. An endpoint that migrates
away from a local address SHOULD retire all connection IDs used on away from a local address SHOULD retire all connection IDs used on
that address once it no longer plans to use that address. that address once it no longer plans to use that address.
6.2. Matching Packets to Connections 5.2. Matching Packets to Connections
Incoming packets are classified on receipt. Packets can either be Incoming packets are classified on receipt. Packets can either be
associated with an existing connection, or - for servers - associated with an existing connection, or - for servers -
potentially create a new connection. potentially create a new connection.
Hosts try to associate a packet with an existing connection. If the Hosts try to associate a packet with an existing connection. If the
packet has a Destination Connection ID corresponding to an existing packet has a Destination Connection ID corresponding to an existing
connection, QUIC processes that packet accordingly. Note that more connection, QUIC processes that packet accordingly. Note that more
than one connection ID can be associated with a connection; see than one connection ID can be associated with a connection; see
Section 6.1. Section 5.1.
If the Destination Connection ID is zero length and the packet If the Destination Connection ID is zero length and the packet
matches the address/port tuple of a connection where the host did not matches the address/port tuple of a connection where the host did not
require connection IDs, QUIC processes the packet as part of that require connection IDs, QUIC processes the packet as part of that
connection. Endpoints MUST drop packets with zero-length Destination connection. Endpoints MUST drop packets with zero-length Destination
Connection ID fields if they do not correspond to a single Connection ID fields if they do not correspond to a single
connection. connection.
Endpoints SHOULD send a Stateless Reset (Section 6.13.4) for any Endpoints SHOULD send a Stateless Reset (Section 10.4) for any
packets that cannot be attributed to an existing connection. packets that cannot be attributed to an existing connection.
6.2.1. Client Packet Handling Packets that are matched to an existing connection, but for which the
endpoint cannot remove packet protection, are discarded.
5.2.1. Client Packet Handling
Valid packets sent to clients always include a Destination Connection Valid packets sent to clients always include a Destination Connection
ID that matches the value the client selects. Clients that choose to ID that matches a value the client selects. Clients that choose to
receive zero-length connection IDs can use the address/port tuple to receive zero-length connection IDs can use the address/port tuple to
identify a connection. Packets that don't match an existing identify a connection. Packets that don't match an existing
connection are discarded. connection are discarded.
Due to packet reordering or loss, clients might receive packets for a Due to packet reordering or loss, clients might receive packets for a
connection that are encrypted with a key it has not yet computed. connection that are encrypted with a key it has not yet computed.
Clients MAY drop these packets, or MAY buffer them in anticipation of Clients MAY drop these packets, or MAY buffer them in anticipation of
later packets that allow it to compute the key. later packets that allow it to compute the key.
If a client receives a packet that has an unsupported version, it If a client receives a packet that has an unsupported version, it
MUST discard that packet. MUST discard that packet.
6.2.2. Server Packet Handling 5.2.2. Server Packet Handling
If a server receives a packet that has an unsupported version, but If a server receives a packet that has an unsupported version, but
the packet is sufficiently large to initiate a new connection for any the packet is sufficiently large to initiate a new connection for any
version supported by the server, it SHOULD send a Version Negotiation version supported by the server, it SHOULD send a Version Negotiation
packet as described in Section 6.3.1. Servers MAY rate control these packet as described in Section 6.1. Servers MAY rate control these
packets to avoid storms of Version Negotiation packets. packets to avoid storms of Version Negotiation packets.
The first packet for an unsupported version can use different The first packet for an unsupported version can use different
semantics and encodings for any version-specific field. In semantics and encodings for any version-specific field. In
particular, different packet protection keys might be used for particular, different packet protection keys might be used for
different versions. Servers that do not support a particular version different versions. Servers that do not support a particular version
are unlikely to be able to decrypt the payload of the packet. are unlikely to be able to decrypt the payload of the packet.
Servers SHOULD NOT attempt to decode or decrypt a packet from an Servers SHOULD NOT attempt to decode or decrypt a packet from an
unknown version, but instead send a Version Negotiation packet, unknown version, but instead send a Version Negotiation packet,
provided that the packet is sufficiently long. provided that the packet is sufficiently long.
Servers MUST drop other packets that contain unsupported versions. Servers MUST drop other packets that contain unsupported versions.
Packets with a supported version, or no version field, are matched to Packets with a supported version, or no version field, are matched to
a connection as described in Section 6.2. If not matched, the server a connection using the connection ID or - for packets with zero-
continues below. length connection IDs - the address tuple. If the packet doesn't
match an existing connection, the server continues below.
If the packet is an Initial packet fully conforming with the If the packet is an Initial packet fully conforming with the
specification, the server proceeds with the handshake (Section 6.4). specification, the server proceeds with the handshake (Section 7).
This commits the server to the version that the client selected. This commits the server to the version that the client selected.
If a server isn't currently accepting any new connections, it SHOULD If a server isn't currently accepting any new connections, it SHOULD
send an Initial packet containing a CONNECTION_CLOSE frame with error send an Initial packet containing a CONNECTION_CLOSE frame with error
code SERVER_BUSY. code SERVER_BUSY.
If the packet is a 0-RTT packet, the server MAY buffer a limited If the packet is a 0-RTT packet, the server MAY buffer a limited
number of these packets in anticipation of a late-arriving Initial number of these packets in anticipation of a late-arriving Initial
Packet. Clients are forbidden from sending Handshake packets prior Packet. Clients are forbidden from sending Handshake packets prior
to receiving a server response, so servers SHOULD ignore any such to receiving a server response, so servers SHOULD ignore any such
packets. packets.
Servers MUST drop incoming packets under all other circumstances. Servers MUST drop incoming packets under all other circumstances.
6.3. Version Negotiation 5.3. Life of a QUIC Connection
TBD.
6. Version Negotiation
Version negotiation ensures that client and server agree to a QUIC Version negotiation ensures that client and server agree to a QUIC
version that is mutually supported. A server sends a Version version that is mutually supported. A server sends a Version
Negotiation packet in response to each packet that might initiate a Negotiation packet in response to each packet that might initiate a
new connection, see Section 6.2 for details. new connection, see Section 5.2 for details.
The first few messages of an exchange between a client attempting to
create a new connection with server is shown in Figure 3. After
version negotiation completes, connection establishment can proceed,
for example as shown in Section 7.1.
Client Server
Packet (v=X) ->
<- Version Negotiation (supported=Y,Z)
Packet (v=Y) ->
<- Packet(s) (v=Y)
Figure 3: Example Version Negotiation Exchange
The size of the first packet sent by a client will determine whether The size of the first packet sent by a client will determine whether
a server sends a Version Negotiation packet. Clients that support a server sends a Version Negotiation packet. Clients that support
multiple QUIC versions SHOULD pad the first packet they send to the multiple QUIC versions SHOULD pad the first packet they send to the
largest of the minimum packet sizes across all versions they support. largest of the minimum packet sizes across all versions they support.
This ensures that the server responds if there is a mutually This ensures that the server responds if there is a mutually
supported version. supported version.
6.3.1. Sending Version Negotiation Packets 6.1. Sending Version Negotiation Packets
If the version selected by the client is not acceptable to the If the version selected by the client is not acceptable to the
server, the server responds with a Version Negotiation packet (see server, the server responds with a Version Negotiation packet (see
Section 4.3). This includes a list of versions that the server will Section 17.4). This includes a list of versions that the server will
accept. accept.
This system allows a server to process packets with unsupported This system allows a server to process packets with unsupported
versions without retaining state. Though either the Initial packet versions without retaining state. Though either the Initial packet
or the Version Negotiation packet that is sent in response could be or the Version Negotiation packet that is sent in response could be
lost, the client will send new packets until it successfully receives lost, the client will send new packets until it successfully receives
a response or it abandons the connection attempt. a response or it abandons the connection attempt.
6.3.2. Handling Version Negotiation Packets A server MAY limit the number of Version Negotiation packets it
sends. For instance, a server that is able to recognize packets as
0-RTT might choose not to send Version Negotiation packets in
response to 0-RTT packets with the expectation that it will
eventually receive an Initial packet.
6.2. Handling Version Negotiation Packets
When the client receives a Version Negotiation packet, it first When the client receives a Version Negotiation packet, it first
checks that the Destination and Source Connection ID fields match the checks that the Destination and Source Connection ID fields match the
Source and Destination Connection ID fields in a packet that the Source and Destination Connection ID fields in a packet that the
client sent. If this check fails, the packet MUST be discarded. client sent. If this check fails, the packet MUST be discarded.
Once the Version Negotiation packet is determined to be valid, the Once the Version Negotiation packet is determined to be valid, the
client then selects an acceptable protocol version from the list client then selects an acceptable protocol version from the list
provided by the server. The client then attempts to create a provided by the server. The client then attempts to create a
connection using that version. Though the content of the Initial connection using that version. Though the content of the Initial
packet the client sends might not change in response to version packet the client sends might not change in response to version
negotiation, a client MUST increase the packet number it uses on negotiation, a client MUST increase the packet number it uses on
every packet it sends. Packets MUST continue to use long headers and every packet it sends. Packets MUST continue to use long headers
MUST include the new negotiated protocol version. (Section 17.2) and MUST include the new negotiated protocol version.
The client MUST use the long header format and include its selected The client MUST use the long header format and include its selected
version on all packets until it has 1-RTT keys and it has received a version on all packets until it has 1-RTT keys and it has received a
packet from the server which is not a Version Negotiation packet. packet from the server which is not a Version Negotiation packet.
A client MUST NOT change the version it uses unless it is in response A client MUST NOT change the version it uses unless it is in response
to a Version Negotiation packet from the server. Once a client to a Version Negotiation packet from the server. Once a client
receives a packet from the server which is not a Version Negotiation receives a packet from the server which is not a Version Negotiation
packet, it MUST discard other Version Negotiation packets on the same packet, it MUST discard other Version Negotiation packets on the same
connection. Similarly, a client MUST ignore a Version Negotiation connection. Similarly, a client MUST ignore a Version Negotiation
packet if it has already received and acted on a Version Negotiation packet if it has already received and acted on a Version Negotiation
packet. packet.
A client MUST ignore a Version Negotiation packet that lists the A client MUST ignore a Version Negotiation packet that lists the
client's chosen version. client's chosen version.
A client MAY attempt 0-RTT after receiving a Version Negotiation A client MAY attempt 0-RTT after receiving a Version Negotiation
packet. A client that sends additional 0-RTT packets MUST NOT reset packet. A client that sends additional 0-RTT packets MUST NOT reset
the packet number to 0 as a result, see Section 4.6.4. the packet number to 0 as a result, see Section 17.5.2.
Version negotiation packets have no cryptographic protection. The Version negotiation packets have no cryptographic protection. The
result of the negotiation MUST be revalidated as part of the result of the negotiation MUST be revalidated as part of the
cryptographic handshake (see Section 6.6.4). cryptographic handshake (see Section 7.3.3).
6.3.3. Using Reserved Versions 6.3. Using Reserved Versions
For a server to use a new version in the future, clients must For a server to use a new version in the future, clients must
correctly handle unsupported versions. To help ensure this, a server correctly handle unsupported versions. To help ensure this, a server
SHOULD include a reserved version (see Section 3) while generating a SHOULD include a reserved version (see Section 15) while generating a
Version Negotiation packet. Version Negotiation packet.
The design of version negotiation permits a server to avoid The design of version negotiation permits a server to avoid
maintaining state for packets that it rejects in this fashion. The maintaining state for packets that it rejects in this fashion. The
validation of version negotiation (see Section 6.6.4) only validates validation of version negotiation (see Section 7.3.3) only validates
the result of version negotiation, which is the same no matter which the result of version negotiation, which is the same no matter which
reserved version was sent. A server MAY therefore send different reserved version was sent. A server MAY therefore send different
reserved version numbers in the Version Negotiation Packet and in its reserved version numbers in the Version Negotiation Packet and in its
transport parameters. transport parameters.
A client MAY send a packet using a reserved version number. This can A client MAY send a packet using a reserved version number. This can
be used to solicit a list of supported versions from a server. be used to solicit a list of supported versions from a server.
6.4. Cryptographic and Transport Handshake 7. Cryptographic and Transport Handshake
QUIC relies on a combined cryptographic and transport handshake to QUIC relies on a combined cryptographic and transport handshake to
minimize connection establishment latency. QUIC uses the CRYPTO minimize connection establishment latency. QUIC uses the CRYPTO
frame Section 7.21 to transmit the cryptographic handshake. Version frame Section 19.20 to transmit the cryptographic handshake. Version
0x00000001 of QUIC uses TLS 1.3 as described in [QUIC-TLS]; a 0x00000001 of QUIC uses TLS 1.3 as described in [QUIC-TLS]; a
different QUIC version number could indicate that a different different QUIC version number could indicate that a different
cryptographic handshake protocol is in use. cryptographic handshake protocol is in use.
QUIC provides reliable, ordered delivery of the cryptographic QUIC provides reliable, ordered delivery of the cryptographic
handshake data. QUIC packet protection ensures confidentiality and handshake data. QUIC packet protection ensures confidentiality and
integrity protection that meets the requirements of the cryptographic integrity protection that meets the requirements of the cryptographic
handshake protocol: handshake protocol:
o authenticated key exchange, where o authenticated key exchange, where
skipping to change at page 36, line 41 skipping to change at page 31, line 33
* a client is optionally authenticated, * a client is optionally authenticated,
* every connection produces distinct and unrelated keys, * every connection produces distinct and unrelated keys,
* keying material is usable for packet protection for both 0-RTT * keying material is usable for packet protection for both 0-RTT
and 1-RTT packets, and and 1-RTT packets, and
* 1-RTT keys have forward secrecy * 1-RTT keys have forward secrecy
o authenticated values for the transport parameters of the peer (see o authenticated values for the transport parameters of the peer (see
Section 6.6) Section 7.3)
o authenticated confirmation of version negotiation (see o authenticated confirmation of version negotiation (see
Section 6.6.4) Section 7.3.3)
o authenticated negotiation of an application protocol (TLS uses o authenticated negotiation of an application protocol (TLS uses
ALPN [RFC7301] for this purpose) ALPN [RFC7301] for this purpose)
o for the server, the ability to carry data that provides assurance The first CRYPTO frame from a client MUST be sent in a single packet.
that the client can receive packets that are addressed with the Any second attempt that is triggered by address validation (see
transport address that is claimed by the client (see Section 6.9) Section 8.1) MUST also be sent within a single packet. This avoids
having to reassemble a message from multiple packets.
The first CRYPTO frame MUST be sent in a single packet. Any second
attempt that is triggered by address validation MUST also be sent
within a single packet. This avoids having to reassemble a message
from multiple packets.
The first client packet of the cryptographic handshake protocol MUST The first client packet of the cryptographic handshake protocol MUST
fit within a 1232 octet QUIC packet payload. This includes overheads fit within a 1232 octet QUIC packet payload. This includes overheads
that reduce the space available to the cryptographic handshake that reduce the space available to the cryptographic handshake
protocol. protocol.
The CRYPTO frame can be sent in different packet number spaces. The CRYPTO frame can be sent in different packet number spaces. The
CRYPTO frames in each packet number space carry a separate sequence sequence numbers used by CRYPTO frames to ensure ordered delivery of
of handshake data starting from an offset of 0. cryptographic handshake data start from zero in each packet number
space.
6.5. Example Handshake Flows 7.1. Example Handshake Flows
Details of how TLS is integrated with QUIC are provided in Details of how TLS is integrated with QUIC are provided in
[QUIC-TLS], but some examples are provided here. [QUIC-TLS], but some examples are provided here. An extension of
this exchange to support client address validation is shown in
Section 8.1.1.
Figure 8 provides an overview of the 1-RTT handshake. Each line Once any version negotiation and address validation exchanges are
complete, the cryptographic handshake is used to agree on
cryptographic keys. The cryptographic handshake is carried in
Initial (Section 17.5) and Handshake (Section 17.6) packets.
Figure 4 provides an overview of the 1-RTT handshake. Each line
shows a QUIC packet with the packet type and packet number shown shows a QUIC packet with the packet type and packet number shown
first, followed by the frames that are typically contained in those first, followed by the frames that are typically contained in those
packets. So, for instance the first packet is of type Initial, with packets. So, for instance the first packet is of type Initial, with
packet number 0, and contains a CRYPTO frame carrying the packet number 0, and contains a CRYPTO frame carrying the
ClientHello. ClientHello.
Note that multiple QUIC packets - even of different encryption levels Note that multiple QUIC packets - even of different encryption levels
- may be coalesced into a single UDP datagram (see Section 4.9), and - may be coalesced into a single UDP datagram (see Section 12.2), and
so this handshake may consist of as few as 4 UDP datagrams, or any so this handshake may consist of as few as 4 UDP datagrams, or any
number more. For instance, the server's first flight contains number more. For instance, the server's first flight contains
packets from the Initial encryption level (obfuscation), the packets from the Initial encryption level (obfuscation), the
Handshake level, and "0.5-RTT data" from the server at the 1-RTT Handshake level, and "0.5-RTT data" from the server at the 1-RTT
encryption level. encryption level.
Client Server Client Server
Initial[0]: CRYPTO[CH] -> Initial[0]: CRYPTO[CH] ->
skipping to change at page 38, line 20 skipping to change at page 32, line 49
Handshake[0]: CRYPTO[EE, CERT, CV, FIN] Handshake[0]: CRYPTO[EE, CERT, CV, FIN]
<- 1-RTT[0]: STREAM[1, "..."] <- 1-RTT[0]: STREAM[1, "..."]
Initial[1]: ACK[0] Initial[1]: ACK[0]
Handshake[0]: CRYPTO[FIN], ACK[0] Handshake[0]: CRYPTO[FIN], ACK[0]
1-RTT[0]: STREAM[0, "..."], ACK[0] -> 1-RTT[0]: STREAM[0, "..."], ACK[0] ->
1-RTT[1]: STREAM[55, "..."], ACK[0] 1-RTT[1]: STREAM[55, "..."], ACK[0]
<- Handshake[1]: ACK[0] <- Handshake[1]: ACK[0]
Figure 8: Example 1-RTT Handshake Figure 4: Example 1-RTT Handshake
Figure 9 shows an example of a connection with a 0-RTT handshake and Figure 5 shows an example of a connection with a 0-RTT handshake and
a single packet of 0-RTT data. Note that as described in a single packet of 0-RTT data. Note that as described in
Section 4.11, the server ACKs the 0-RTT data at the 1-RTT encryption Section 12.3, the server acknowledges 0-RTT data at the 1-RTT
level, and the client's sequence numbers at the 1-RTT encryption encryption level, and the client sends 1-RTT packets in the same
level continue to increment from its 0-RTT packets. packet number space.
Client Server Client Server
Initial[0]: CRYPTO[CH] Initial[0]: CRYPTO[CH]
0-RTT[0]: STREAM[0, "..."] -> 0-RTT[0]: STREAM[0, "..."] ->
Initial[0]: CRYPTO[SH] ACK[0] Initial[0]: CRYPTO[SH] ACK[0]
Handshake[0] CRYPTO[EE, CERT, CV, FIN] Handshake[0] CRYPTO[EE, CERT, CV, FIN]
<- 1-RTT[0]: STREAM[1, "..."] ACK[0] <- 1-RTT[0]: STREAM[1, "..."] ACK[0]
Initial[1]: ACK[0] Initial[1]: ACK[0]
0-RTT[1]: CRYPTO[EOED]
Handshake[0]: CRYPTO[FIN], ACK[0] Handshake[0]: CRYPTO[FIN], ACK[0]
1-RTT[2]: STREAM[0, "..."] ACK[0] -> 1-RTT[2]: STREAM[0, "..."] ACK[0] ->
1-RTT[1]: STREAM[55, "..."], ACK[1,2] 1-RTT[1]: STREAM[55, "..."], ACK[1,2]
<- Handshake[1]: ACK[0] <- Handshake[1]: ACK[0]
Figure 9: Example 0-RTT Handshake Figure 5: Example 0-RTT Handshake
6.6. Transport Parameters 7.2. Negotiating Connection IDs
During connection establishment, both endpoints make authenticated A connection ID is used to ensure consistent routing of packets, as
declarations of their transport parameters. These declarations are described in Section 5.1. The long header contains two connection
made unilaterally by each endpoint. Endpoints are required to comply IDs: the Destination Connection ID is chosen by the recipient of the
with the restrictions implied by these parameters; the description of packet and is used to provide consistent routing; the Source
each parameter includes rules for its handling. Connection ID is used to set the Destination Connection ID used by
the peer.
The format of the transport parameters is the TransportParameters During the handshake, packets with the long header (Section 17.2) are
struct from Figure 10. This is described using the presentation used to establish the connection ID that each endpoint uses. Each
language from Section 3 of [TLS13]. endpoint uses the Source Connection ID field to specify the
connection ID that is used in the Destination Connection ID field of
packets being sent to them. Upon receiving a packet, each endpoint
sets the Destination Connection ID it sends to match the value of the
Source Connection ID that they receive.
uint32 QuicVersion; When an Initial packet is sent by a client which has not previously
received a Retry packet from the server, it populates the Destination
Connection ID field with an unpredictable value. This MUST be at
least 8 octets in length. Until a packet is received from the
server, the client MUST use the same value unless it abandons the
connection attempt and starts a new one. The initial Destination
Connection ID is used to determine packet protection keys for Initial
packets.
enum { The client populates the Source Connection ID field with a value of
initial_max_stream_data_bidi_local(0), its choosing and sets the SCIL field to match.
initial_max_data(1),
initial_max_bidi_streams(2),
idle_timeout(3),
preferred_address(4),
max_packet_size(5),
stateless_reset_token(6),
ack_delay_exponent(7),
initial_max_uni_streams(8),
disable_migration(9),
initial_max_stream_data_bidi_remote(10),
initial_max_stream_data_uni(11),
max_ack_delay(12),
original_connection_id(13),
(65535)
} TransportParameterId;
struct { The Destination Connection ID field in the server's Initial packet
TransportParameterId parameter; contains a connection ID that is chosen by the recipient of the
opaque value<0..2^16-1>; packet (i.e., the client); the Source Connection ID includes the
} TransportParameter; connection ID that the sender of the packet wishes to use (see
Section 5.1). The server MUST use consistent Source Connection IDs
during the handshake.
struct { On first receiving an Initial or Retry packet from the server, the
select (Handshake.msg_type) { client uses the Source Connection ID supplied by the server as the
case client_hello: Destination Connection ID for subsequent packets. That means that a
QuicVersion initial_version; client might change the Destination Connection ID twice during
connection establishment. Once a client has received an Initial
packet from the server, it MUST discard any packet it receives with a
different Source Connection ID.
case encrypted_extensions: A client MUST only change the value it sends in the Destination
QuicVersion negotiated_version; Connection ID in response to the first packet of each type it
QuicVersion supported_versions<4..2^8-4>; receives from the server (Retry or Initial); a server MUST set its
}; value based on the Initial packet. Any additional changes are not
TransportParameter parameters<22..2^16-1>; permitted; if subsequent packets of those types include a different
} TransportParameters; Source Connection ID, they MUST be discarded. This avoids problems
that might arise from stateless processing of multiple Initial
packets producing different connection IDs.
struct { The connection ID can change over the lifetime of a connection,
enum { IPv4(4), IPv6(6), (15) } ipVersion; especially in response to connection migration (Section 9), see
opaque ipAddress<4..2^8-1>; Section 5.1.1 for details.
uint16 port;
opaque connectionId<0..18>;
opaque statelessResetToken[16];
} PreferredAddress;
Figure 10: Definition of TransportParameters 7.3. Transport Parameters
The "extension_data" field of the quic_transport_parameters extension During connection establishment, both endpoints make authenticated
defined in [QUIC-TLS] contains a TransportParameters value. TLS declarations of their transport parameters. These declarations are
encoding rules are therefore used to encode the transport parameters. made unilaterally by each endpoint. Endpoints are required to comply
with the restrictions implied by these parameters; the description of
each parameter includes rules for its handling.
QUIC encodes transport parameters into a sequence of octets, which The encoding of the transport parameters is detailed in Section 18.
are then included in the cryptographic handshake. Once the handshake
completes, the transport parameters declared by the peer are QUIC includes the encoded transport parameters in the cryptographic
available. Each endpoint validates the value provided by its peer. handshake. Once the handshake completes, the transport parameters
In particular, version negotiation MUST be validated (see declared by the peer are available. Each endpoint validates the
Section 6.6.4) before the connection establishment is considered value provided by its peer. In particular, version negotiation MUST
properly complete. be validated (see Section 7.3.3) before the connection establishment
is considered properly complete.
Definitions for each of the defined transport parameters are included Definitions for each of the defined transport parameters are included
in Section 6.6.1. Any given parameter MUST appear at most once in a in Section 18.1. Any given parameter MUST appear at most once in a
given transport parameters extension. An endpoint MUST treat receipt given transport parameters extension. An endpoint MUST treat receipt
of duplicate transport parameters as a connection error of type of duplicate transport parameters as a connection error of type
TRANSPORT_PARAMETER_ERROR. TRANSPORT_PARAMETER_ERROR.
6.6.1. Transport Parameter Definitions
An endpoint MAY use the following transport parameters:
initial_max_data (0x0001): The initial maximum data parameter
contains the initial value for the maximum amount of data that can
be sent on the connection. This parameter is encoded as an
unsigned 32-bit integer in units of octets. This is equivalent to
sending a MAX_DATA (Section 7.6) for the connection immediately
after completing the handshake. If the transport parameter is
absent, the connection starts with a flow control limit of 0.
initial_max_bidi_streams (0x0002): The initial maximum bidirectional
streams parameter contains the initial maximum number of
bidirectional streams the peer may initiate, encoded as an
unsigned 16-bit integer. If this parameter is absent or zero,
bidirectional streams cannot be created until a MAX_STREAM_ID
frame is sent. Setting this parameter is equivalent to sending a
MAX_STREAM_ID (Section 7.8) immediately after completing the
handshake containing the corresponding Stream ID. For example, a
value of 0x05 would be equivalent to receiving a MAX_STREAM_ID
containing 16 when received by a client or 17 when received by a
server.
initial_max_uni_streams (0x0008): The initial maximum unidirectional
streams parameter contains the initial maximum number of
unidirectional streams the peer may initiate, encoded as an
unsigned 16-bit integer. If this parameter is absent or zero,
unidirectional streams cannot be created until a MAX_STREAM_ID
frame is sent. Setting this parameter is equivalent to sending a
MAX_STREAM_ID (Section 7.8) immediately after completing the
handshake containing the corresponding Stream ID. For example, a
value of 0x05 would be equivalent to receiving a MAX_STREAM_ID
containing 18 when received by a client or 19 when received by a
server.
idle_timeout (0x0003): The idle timeout is a value in seconds that
is encoded as an unsigned 16-bit integer. If this parameter is
absent or zero then the idle timeout is disabled.
max_packet_size (0x0005): The maximum packet size parameter places a
limit on the size of packets that the endpoint is willing to
receive, encoded as an unsigned 16-bit integer. This indicates
that packets larger than this limit will be dropped. The default
for this parameter is the maximum permitted UDP payload of 65527.
Values below 1200 are invalid. This limit only applies to
protected packets (Section 4.8).
ack_delay_exponent (0x0007): An 8-bit unsigned integer value
indicating an exponent used to decode the ACK Delay field in the
ACK frame, see Section 7.16. If this value is absent, a default
value of 3 is assumed (indicating a multiplier of 8). The default
value is also used for ACK frames that are sent in Initial and
Handshake packets. Values above 20 are invalid.
disable_migration (0x0009): The endpoint does not support connection
migration (Section 6.11). Peers MUST NOT send any packets,
including probing packets (Section 6.11.1), from a local address
other than that used to perform the handshake. This parameter is
a zero-length value.
max_ack_delay (0x000c): An 8 bit unsigned integer value indicating
the maximum amount of time in milliseconds by which it will delay
sending of acknowledgments. If this value is absent, a default of
25 milliseconds is assumed.
Either peer MAY advertise an initial value for the flow control on
each type of stream on which they might receive data. Each of the
following transport parameters is encoded as an unsigned 32-bit
integer in units of octets:
initial_max_stream_data_bidi_local (0x0000): The initial stream
maximum data for bidirectional, locally-initiated streams
parameter contains the initial flow control limit for newly
created bidirectional streams opened by the endpoint that sets the
transport parameter. In client transport parameters, this applies
to streams with an identifier ending in 0x0; in server transport
parameters, this applies to streams ending in 0x1.
initial_max_stream_data_bidi_remote (0x000a): The initial stream
maximum data for bidirectional, peer-initiated streams parameter
contains the initial flow control limit for newly created
bidirectional streams opened by the endpoint that receives the
transport parameter. In client transport parameters, this applies
to streams with an identifier ending in 0x1; in server transport
parameters, this applies to streams ending in 0x0.
initial_max_stream_data_uni (0x000b): The initial stream maximum
data for unidirectional streams parameter contains the initial
flow control limit for newly created unidirectional streams opened
by the endpoint that receives the transport parameter. In client
transport parameters, this applies to streams with an identifier
ending in 0x3; in server transport parameters, this applies to
streams ending in 0x2.
If present, transport parameters that set initial stream flow control
limits are equivalent to sending a MAX_STREAM_DATA frame
(Section 7.7) on every stream of the corresponding type immediately
after opening. If the transport parameter is absent, streams of that
type start with a flow control limit of 0.
A server MUST include the original_connection_id transport parameter A server MUST include the original_connection_id transport parameter
if it sent a Retry packet: (Section 18.1) if it sent a Retry packet.
original_connection_id (0x000d): The value of the Destination
Connection ID field from the first Initial packet sent by the
client. This transport parameter is only sent by the server.
A server MAY include the following transport parameters:
stateless_reset_token (0x0006): The Stateless Reset Token is used in
verifying a stateless reset, see Section 6.13.4. This parameter
is a sequence of 16 octets.
preferred_address (0x0004): The server's Preferred Address is used
to effect a change in server address at the end of the handshake,
as described in Section 6.12.
A client MUST NOT include a stateless reset token or a preferred
address. A server MUST treat receipt of either transport parameter
as a connection error of type TRANSPORT_PARAMETER_ERROR.
6.6.2. Values of Transport Parameters for 0-RTT 7.3.1. Values of Transport Parameters for 0-RTT
A client that attempts to send 0-RTT data MUST remember the transport A client that attempts to send 0-RTT data MUST remember the transport
parameters used by the server. The transport parameters that the parameters used by the server. The transport parameters that the
server advertises during connection establishment apply to all server advertises during connection establishment apply to all
connections that are resumed using the keying material established connections that are resumed using the keying material established
during that handshake. Remembered transport parameters apply to the during that handshake. Remembered transport parameters apply to the
new connection until the handshake completes and new transport new connection until the handshake completes and new transport
parameters from the server can be provided. parameters from the server can be provided.
A server can remember the transport parameters that it advertised, or A server can remember the transport parameters that it advertised, or
store an integrity-protected copy of the values in the ticket and store an integrity-protected copy of the values in the ticket and
recover the information when accepting 0-RTT data. A server uses the recover the information when accepting 0-RTT data. A server uses the
transport parameters in determining whether to accept 0-RTT data. transport parameters in determining whether to accept 0-RTT data.
A server MAY accept 0-RTT and subsequently provide different values A server MAY accept 0-RTT and subsequently provide different values
for transport parameters for use in the new connection. If 0-RTT for transport parameters for use in the new connection. If 0-RTT
data is accepted by the server, the server MUST NOT reduce any limits data is accepted by the server, the server MUST NOT reduce any limits
or alter any values that might be violated by the client with its or alter any values that might be violated by the client with its
0-RTT data. In particular, a server that accepts 0-RTT data MUST NOT 0-RTT data. In particular, a server that accepts 0-RTT data MUST NOT
set values for initial_max_data, initial_max_stream_data_bidi_local, set values for initial_max_data, initial_max_stream_data_bidi_local,
initial_max_stream_data_bidi_remote, and initial_max_stream_data_uni initial_max_stream_data_bidi_remote, initial_max_stream_data_uni,
initial_max_bidi_streams, or initial_max_uni_streams (Section 18.1)
that are smaller than the remembered value of those parameters. that are smaller than the remembered value of those parameters.
Similarly, a server MUST NOT reduce the value of
initial_max_bidi_streams or initial_max_uni_streams.
Omitting or setting a zero value for certain transport parameters can Omitting or setting a zero value for certain transport parameters can
result in 0-RTT data being enabled, but not usable. The applicable result in 0-RTT data being enabled, but not usable. The applicable
subset of transport parameters that permit sending of application subset of transport parameters that permit sending of application
data SHOULD be set to non-zero values for 0-RTT. This includes data SHOULD be set to non-zero values for 0-RTT. This includes
initial_max_data and either initial_max_bidi_streams and initial_max_data and either initial_max_bidi_streams and
initial_max_stream_data_bidi_remote, or initial_max_uni_streams and initial_max_stream_data_bidi_remote, or initial_max_uni_streams and
initial_max_stream_data_uni. initial_max_stream_data_uni.
The value of the server's previous preferred_address MUST NOT be used The value of the server's previous preferred_address MUST NOT be used
when establishing a new connection; rather, the client should wait to when establishing a new connection; rather, the client should wait to
observe the server's new preferred_address value in the handshake. observe the server's new preferred_address value in the handshake.
A server MUST reject 0-RTT data or even abort a handshake if the A server MUST reject 0-RTT data or even abort a handshake if the
implied values for transport parameters cannot be supported. implied values for transport parameters cannot be supported.
6.6.3. New Transport Parameters 7.3.2. New Transport Parameters
New transport parameters can be used to negotiate new protocol New transport parameters can be used to negotiate new protocol
behavior. An endpoint MUST ignore transport parameters that it does behavior. An endpoint MUST ignore transport parameters that it does
not support. Absence of a transport parameter therefore disables any not support. Absence of a transport parameter therefore disables any
optional protocol feature that is negotiated using the parameter. optional protocol feature that is negotiated using the parameter.
New transport parameters can be registered according to the rules in New transport parameters can be registered according to the rules in
Section 13.1. Section 22.1.
6.6.4. Version Negotiation Validation 7.3.3. Version Negotiation Validation
Though the cryptographic handshake has integrity protection, two Though the cryptographic handshake has integrity protection, two
forms of QUIC version downgrade are possible. In the first, an forms of QUIC version downgrade are possible. In the first, an
attacker replaces the QUIC version in the Initial packet. In the attacker replaces the QUIC version in the Initial packet. In the
second, a fake Version Negotiation packet is sent by an attacker. To second, a fake Version Negotiation packet is sent by an attacker. To
protect against these attacks, the transport parameters include three protect against these attacks, the transport parameters include three
fields that encode version information. These parameters are used to fields that encode version information. These parameters are used to
retroactively authenticate the choice of version (see Section 6.3). retroactively authenticate the choice of version (see Section 6).
The cryptographic handshake provides integrity protection for the The cryptographic handshake provides integrity protection for the
negotiated version as part of the transport parameters (see negotiated version as part of the transport parameters (see
Section 6.6). As a result, attacks on version negotiation by an Section 18.1). As a result, attacks on version negotiation by an
attacker can be detected. attacker can be detected.
The client includes the initial_version field in its transport The client includes the initial_version field in its transport
parameters. The initial_version is the version that the client parameters. The initial_version is the version that the client
initially attempted to use. If the server did not send a Version initially attempted to use. If the server did not send a Version
Negotiation packet Section 4.3, this will be identical to the Negotiation packet Section 17.4, this will be identical to the
negotiated_version field in the server transport parameters. negotiated_version field in the server transport parameters.
A server that processes all packets in a stateful fashion can A server that processes all packets in a stateful fashion can
remember how version negotiation was performed and validate the remember how version negotiation was performed and validate the
initial_version value. initial_version value.
A server that does not maintain state for every packet it receives A server that does not maintain state for every packet it receives
(i.e., a stateless server) uses a different process. If the (i.e., a stateless server) uses a different process. If the
initial_version matches the version of QUIC that is in use, a initial_version matches the version of QUIC that is in use, a
stateless server can accept the value. stateless server can accept the value.
If the initial_version is different from the version of QUIC that is If the initial_version is different from the version of QUIC that is
in use, a stateless server MUST check that it would have sent a in use, a stateless server MUST check that it would have sent a
Version Negotiation packet if it had received a packet with the Version Negotiation packet if it had received a packet with the
indicated initial_version. If a server would have accepted the indicated initial_version. If a server would have accepted the
version included in the initial_version and the value differs from version included in the initial_version and the value differs from
the QUIC version that is in use, the server MUST terminate the the QUIC version that is in use, the server MUST terminate the
connection with a VERSION_NEGOTIATION_ERROR error. connection with a VERSION_NEGOTIATION_ERROR error.
The server includes both the version of QUIC that is in use and a The server includes both the version of QUIC that is in use and a
list of the QUIC versions that the server supports. list of the QUIC versions that the server supports (see
Section 18.1).
The negotiated_version field is the version that is in use. This The negotiated_version field is the version that is in use. This
MUST be set by the server to the value that is on the Initial packet MUST be set by the server to the value that is on the Initial packet
that it accepts (not an Initial packet that triggers a Retry or that it accepts (not an Initial packet that triggers a Retry or
Version Negotiation packet). A client that receives a Version Negotiation packet). A client that receives a
negotiated_version that does not match the version of QUIC that is in negotiated_version that does not match the version of QUIC that is in
use MUST terminate the connection with a VERSION_NEGOTIATION_ERROR use MUST terminate the connection with a VERSION_NEGOTIATION_ERROR
error code. error code.
The server includes a list of versions that it would send in any The server includes a list of versions that it would send in any
version negotiation packet (Section 4.3) in the supported_versions version negotiation packet (Section 17.4) in the supported_versions
field. The server populates this field even if it did not send a field. The server populates this field even if it did not send a
version negotiation packet. version negotiation packet.
The client validates that the negotiated_version is included in the The client validates that the negotiated_version is included in the
supported_versions list and - if version negotiation was performed - supported_versions list and - if version negotiation was performed -
that it would have selected the negotiated version. A client MUST that it would have selected the negotiated version. A client MUST
terminate the connection with a VERSION_NEGOTIATION_ERROR error code terminate the connection with a VERSION_NEGOTIATION_ERROR error code
if the current QUIC version is not listed in the supported_versions if the current QUIC version is not listed in the supported_versions
list. A client MUST terminate with a VERSION_NEGOTIATION_ERROR error list. A client MUST terminate with a VERSION_NEGOTIATION_ERROR error
code if version negotiation occurred but it would have selected a code if version negotiation occurred but it would have selected a
skipping to change at page 46, line 29 skipping to change at page 37, line 43
When an endpoint accepts multiple QUIC versions, it can potentially When an endpoint accepts multiple QUIC versions, it can potentially
interpret transport parameters as they are defined by any of the QUIC interpret transport parameters as they are defined by any of the QUIC
versions it supports. The version field in the QUIC packet header is versions it supports. The version field in the QUIC packet header is
authenticated using transport parameters. The position and the authenticated using transport parameters. The position and the
format of the version fields in transport parameters MUST either be format of the version fields in transport parameters MUST either be
identical across different QUIC versions, or be unambiguously identical across different QUIC versions, or be unambiguously
different to ensure no confusion about their interpretation. One way different to ensure no confusion about their interpretation. One way
that a new format could be introduced is to define a TLS extension that a new format could be introduced is to define a TLS extension
with a different codepoint. with a different codepoint.
6.7. Stateless Retries 8. Address Validation
A server can process an Initial packet from a client without
committing any state. This allows a server to perform address
validation (Section 6.9), or to defer connection establishment costs.
A server that generates a response to an Initial packet without
retaining connection state MUST use the Retry packet (Section 4.4).
This packet causes a client to restart the connection attempt and
includes the token in the new Initial packet (Section 4.6) to prove
source address ownership.
6.8. Using Explicit Congestion Notification
QUIC endpoints use Explicit Congestion Notification (ECN) [RFC3168]
to detect and respond to network congestion. ECN allows a network
node to indicate congestion in the network by setting a codepoint in
the IP header of a packet instead of dropping it. Endpoints react to
congestion by reducing their sending rate in response, as described
in [QUIC-RECOVERY].
To use ECN, QUIC endpoints first determine whether a path supports
ECN marking and the peer is able to access the ECN codepoint in the
IP header. A network path does not support ECN if ECN marked packets
get dropped or ECN markings are rewritten on the path. An endpoint
verifies the path, both during connection establishment and when
migrating to a new path (see Section 6.11).
Each endpoint independently verifies and enables use of ECN by
setting the IP header ECN codepoint to ECN Capable Transport (ECT)
for the path from it to the other peer. Even if ECN is not used on
the path to the peer, the endpoint MUST provide feedback about ECN
markings received (if accessible).
To verify both that a path supports ECN and the peer can provide ECN
feedback, an endpoint MUST set the ECT(0) codepoint in the IP header
of all outgoing packets [RFC8311].
If an ECT codepoint set in the IP header is not corrupted by a
network device, then a received packet contains either the codepoint
sent by the peer or the Congestion Experienced (CE) codepoint set by
a network device that is experiencing congestion.
On receiving a packet with an ECT or CE codepoint, an endpoint that
can access the IP ECN codepoints increases the corresponding ECT(0),
ECT(1), or CE count, and includes these counters in subsequent (see
Section 8.1) ACK frames (see Section 7.16).
A packet detected by a receiver as a duplicate does not affect the
receiver's local ECN codepoint counts; see (Section 12.7) for
relevant security concerns.
If an endpoint receives a packet without an ECT or CE codepoint, it
responds per Section 8.1 with an ACK frame.
If an endpoint does not have access to received ECN codepoints, it
acknowledges received packets per Section 8.1 with an ACK frame.
If a packet sent with an ECT codepoint is newly acknowledged by the
peer in an ACK frame, the endpoint stops setting ECT codepoints in
subsequent packets, with the expectation that either the network or
the peer no longer supports ECN.
To protect the connection from arbitrary corruption of ECN codepoints
by the network, an endpoint verifies the following when an ACK frame
is received:
o The increase in ECT(0) and ECT(1) counters MUST be at least the
number of packets newly acknowledged that were sent with the
corresponding codepoint.
o The total increase in ECT(0), ECT(1), and CE counters reported in Address validation is used by QUIC to avoid being used for a traffic
the ACK frame MUST be at least the total number of packets newly amplification attack. In such an attack, a packet is sent to a
acknowledged in this ACK frame. server with spoofed source address information that identifies a
victim. If a server generates more or larger packets in response to
that packet, the attacker can use the server to send more data toward
the victim than it would be able to send on its own.
An endpoint could miss acknowledgements for a packet when ACK frames The primary defense against amplification attack is verifying that an
are lost. It is therefore possible for the total increase in ECT(0), endpoint is able to receive packets at the transport address that it
ECT(1), and CE counters to be greater than the number of packets claims. Address validation is performed both during connection
acknowledged in an ACK frame. When this happens, the local reference establishment (see Section 8.1) and during connection migration (see
counts MUST be increased to match the counters in the ACK frame. Section 8.2).
Upon successful verification, an endpoint continues to set ECT 8.1. Address Validation During Connection Establishment
codepoints in subsequent packets with the expectation that the path
is ECN-capable.
If verification fails, then the endpoint ceases setting ECT Connection establishment implicitly provides address validation for
codepoints in subsequent packets with the expectation that either the both endpoints. In particular, receipt of a packet protected with
network or the peer does not support ECN. Handshake keys confirms that the client received the Initial packet
from the server. Once the server has successfully processed a
Handshake packet from the client, it can consider the client address
to have been validated.
If an endpoint sets ECT codepoints on outgoing packets and encounters Prior to validating the client address, servers MUST NOT send more
a retransmission timeout due to the absence of acknowledgments from than three times as many bytes as the number of bytes they have
the peer (see [QUIC-RECOVERY]), or if an endpoint has reason to received. This limits the magnitude of any amplification attack that
believe that a network element might be corrupting ECN codepoints, can be mounted using spoofed source addresses.
the endpoint MAY cease setting ECT codepoints in subsequent packets.
Doing so allows the connection to traverse network elements that drop
or corrupt ECN codepoints in the IP header.
6.9. Proof of Source Address Ownership To ensure that the server is not overly constrained by this
restriction, clients MUST send UDP datagrams with at least 1200
octets of payload until the server has completed address validation,
see Section 14.
Transport protocols commonly spend a round trip checking that a In order to prevent a handshake deadlock as a result of the server
client owns the transport address (IP and port) that it claims. being unable to send, clients SHOULD send a packet upon a handshake
Verifying that a client can receive packets sent to its claimed timeout, as described in [QUIC-RECOVERY]. If the client has no data
transport address protects against spoofing of this information by to retransmit and does not have Handshake keys, it SHOULD send an
malicious clients. Initial packet in a UDP datagram of at least 1200 octets. If the
client has Handshake keys, it SHOULD send a Handshake packet.
This technique is used primarily to avoid QUIC from being used for A server might wish to validate the client address before starting
traffic amplification attack. In such an attack, a packet is sent to the cryptographic handshake. Client addresses can be verified using
a server with spoofed source address information that identifies a an address validation token. This token is delivered during
victim. If a server generates more or larger packets in response to connection establishment with a Retry packet (see Section 8.1.1) or
that packet, the attacker can use the server to send more data toward in a previous connection using the NEW_TOKEN frame (see
the victim than it would be able to send on its own. Section 8.1.2).
Several methods are used in QUIC to mitigate this attack. Firstly, 8.1.1. Address Validation using Retry Packets
the initial handshake packet is sent in a UDP datagram that contains
at least 1200 octets of UDP payload. This allows a server to send a
similar amount of data without risking causing an amplification
attack toward an unproven remote address.
A server eventually confirms that a client has received its messages QUIC uses token-based address validation during connection
when the first Handshake-level message is received. This might be establishment. Any time the server wishes to validate a client
insufficient, either because the server wishes to avoid the address, it provides the client with a token. As long as it is not
computational cost of completing the handshake, or it might be that possible for an attacker to generate a valid token for its own
the size of the packets that are sent during the handshake is too address (see Section 8.1.3) and the client is able to return that
large. This is especially important for 0-RTT, where the server token, it proves to the server that it received the token.
might wish to provide application data traffic - such as a response
to a request - in response to the data carried in the early data from
the client.
To send additional data prior to completing the cryptographic Upon receiving the client's Initial packet, the server can request
handshake, the server then needs to validate that the client owns the address validation by sending a Retry packet (Section 17.7)
address that it claims. containing a token. This token is repeated by the client in an
Initial packet after it receives the Retry packet. In response to
receiving a token in an Initial packet, a server can either abort the
connection or permit it to proceed.
Source address validation is therefore performed by the core A server can also use a Retry packet to defer the state and
transport protocol during the establishment of a connection. processing costs of connection establishment. By giving the client a
different connection ID to use, a server can cause the connection to
be routed to a server instance with more resources available for new
connections.
A different type of source address validation is performed after a A flow showing the use of a Retry packet is shown in Figure 6.
connection migration, see Section 6.10.
6.9.1. Client Address Validation Procedure Client Server
QUIC uses token-based address validation. Any time the server wishes Initial[0]: CRYPTO[CH] ->
to validate a client address, it provides the client with a token.
As long as the token's authenticity can be checked (see
Section 6.9.3) and the client is able to return that token, it proves
to the server that it received the token.
Upon receiving the client's Initial packet, the server can request <- Retry+Token
address validation by sending a Retry packet containing a token.
This token is repeated in the client's next Initial packet. Because
the token is consumed by the server that generates it, there is no
need for a single well-defined format. A token could include
information about the claimed client address (IP and port), a
timestamp, and any other supplementary information the server will
need to validate the token in the future.
The Retry packet is sent to the client and a legitimate client will Initial+Token[0]: CRYPTO[CH] ->
respond with an Initial packet containing the token from the Retry
packet when it continues the handshake. In response to receiving the
token, a server can either abort the connection or permit it to
proceed.
A connection MAY be accepted without address validation - or with Initial[0]: CRYPTO[SH] ACK[0]
only limited validation - but a server SHOULD limit the data it sends Handshake[0]: CRYPTO[EE, CERT, CV, FIN]
toward an unvalidated address. Successful completion of the <- 1-RTT[0]: STREAM[1, "..."]
cryptographic handshake implicitly provides proof that the client has
received packets from the server.
The client should allow for additional Retry packets being sent in Figure 6: Example Handshake with Retry
response to Initial packets sent containing a token. There are
several situations in which the server might not be able to use the
previously generated token to validate the client's address and must
send a new Retry. A reasonable limit to the number of tries the
client allows for, before giving up, is 3. That is, the client MUST
echo the address validation token from a new Retry packet up to 3
times. After that, it MAY give up on the connection attempt.
6.9.2. Address Validation for Future Connections 8.1.2. Address Validation for Future Connections
A server MAY provide clients with an address validation token during A server MAY provide clients with an address validation token during
one connection that can be used on a subsequent connection. Address one connection that can be used on a subsequent connection. Address
validation is especially important with 0-RTT because a server validation is especially important with 0-RTT because a server
potentially sends a significant amount of data to a client in potentially sends a significant amount of data to a client in
response to 0-RTT data. response to 0-RTT data.
The server uses the NEW_TOKEN frame Section 7.19 to provide the The server uses the NEW_TOKEN frame Section 19.18 to provide the
client with an address validation token that can be used to validate client with an address validation token that can be used to validate
future connections. The client may then use this token to validate future connections. The client may then use this token to validate
future connections by including it in the Initial packet's header. future connections by including it in the Initial packet's header.
The client MUST NOT use the token provided in a Retry for future The client MUST NOT use the token provided in a Retry for future
connections. connections.
Unlike the token that is created for a Retry packet, there might be Unlike the token that is created for a Retry packet, there might be
some time between when the token is created and when the token is some time between when the token is created and when the token is
subsequently used. Thus, a resumption token SHOULD include an subsequently used. Thus, a resumption token SHOULD include an
expiration time. The server MAY include either an explicit expiration time. The server MAY include either an explicit
expiration time or an issued timestamp and dynamically calculate the expiration time or an issued timestamp and dynamically calculate the
expiration time. It is also unlikely that the client port number is expiration time. It is also unlikely that the client port number is
the same on two different connections; validating the port is the same on two different connections; validating the port is
therefore unlikely to be successful. therefore unlikely to be successful.
6.9.3. Address Validation Token Integrity A resumption token SHOULD be constructed to be easily distinguishable
from tokens that are sent in Retry packets as they are carried in the
same field.
If the client has a token received in a NEW_TOKEN frame on a previous
connection to what it believes to be the same server, it can include
that value in the Token field of its Initial packet.
A token allows a server to correlate activity between the connection
where the token was issued and any connection where it is used.
Clients that want to break continuity of identity with a server MAY
discard tokens provided using the NEW_TOKEN frame. Tokens obtained
in Retry packets MUST NOT be discarded.
A client SHOULD NOT reuse a token. Reusing a token allows
connections to be linked by entities on the network path (see
Section 9.5). A client MUST NOT reuse a token if it believes that
its point of network attachment has changed since the token was last
used; that is, if there is a change in its local IP address or
network interface. A client needs to start the connection process
over if it migrates prior to completing the handshake.
When a server receives an Initial packet with an address validation
token, it SHOULD attempt to validate it. If the token is invalid
then the server SHOULD proceed as if the client did not have a
validated address, including potentially sending a Retry. If the
validation succeeds, the server SHOULD then allow the handshake to
proceed.
Note: The rationale for treating the client as unvalidated rather
than discarding the packet is that the client might have received
the token in a previous connection using the NEW_TOKEN frame, and
if the server has lost state, it might be unable to validate the
token at all, leading to connection failure if the packet is
discarded. A server MAY encode tokens provided with NEW_TOKEN
frames and Retry packets differently, and validate the latter more
strictly.
In a stateless design, a server can use encrypted and authenticated
tokens to pass information to clients that the server can later
recover and use to validate a client address. Tokens are not
integrated into the cryptographic handshake and so they are not
authenticated. For instance, a client might be able to reuse a
token. To avoid attacks that exploit this property, a server can
limit its use of tokens to only the information needed validate
client addresses.
8.1.3. Address Validation Token Integrity
An address validation token MUST be difficult to guess. Including a An address validation token MUST be difficult to guess. Including a
large enough random value in the token would be sufficient, but this large enough random value in the token would be sufficient, but this
depends on the server remembering the value it sends to clients. depends on the server remembering the value it sends to clients.
A token-based scheme allows the server to offload any state A token-based scheme allows the server to offload any state
associated with validation to the client. For this design to work, associated with validation to the client. For this design to work,
the token MUST be covered by integrity protection against the token MUST be covered by integrity protection against
modification or falsification by clients. Without integrity modification or falsification by clients. Without integrity
protection, malicious clients could generate or guess values for protection, malicious clients could generate or guess values for
tokens that would be accepted by the server. Only the server tokens that would be accepted by the server. Only the server
requires access to the integrity protection key for tokens. requires access to the integrity protection key for tokens.
6.10. Path Validation There is no need for a single well-defined format for the token
because the server that generates the token also consumes it. A
token could include information about the claimed client address (IP
and port), a timestamp, and any other supplementary information the
server will need to validate the token in the future.
Path validation is used by an endpoint to verify reachability of a 8.2. Path Validation
peer over a specific path. That is, it tests reachability between a
specific local address and a specific peer address, where an address
is the two-tuple of IP address and port. Path validation tests that
packets can be both sent to and received from a peer.
Path validation is used during connection migration (see Section 6.11 Path validation is used during connection migration (see Section 9
and Section 6.12) by the migrating endpoint to verify reachability of and Section 9.6) by the migrating endpoint to verify reachability of
a peer from a new local address. Path validation is also used by the a peer from a new local address. In path validation, endpoints test
peer to verify that the migrating endpoint is able to receive packets reachability between a specific local address and a specific peer
sent to the its new address. That is, that the packets received from address, where an address is the two-tuple of IP address and port.
the migrating endpoint do not carry a spoofed source address.
Path validation tests that packets can be both sent to and received
from a peer on the path. Importantly, it validates that the packets
received from the migrating endpoint do not carry a spoofed source
address.
Path validation can be used at any time by either endpoint. For Path validation can be used at any time by either endpoint. For
instance, an endpoint might check that a peer is still in possession instance, an endpoint might check that a peer is still in possession
of its address after a period of quiescence. of its address after a period of quiescence.
Path validation is not designed as a NAT traversal mechanism. Though Path validation is not designed as a NAT traversal mechanism. Though
the mechanism described here might be effective for the creation of the mechanism described here might be effective for the creation of
NAT bindings that support NAT traversal, the expectation is that one NAT bindings that support NAT traversal, the expectation is that one
or other peer is able to receive packets without first having sent a or other peer is able to receive packets without first having sent a
packet on that path. Effective NAT traversal needs additional packet on that path. Effective NAT traversal needs additional
synchronization mechanisms that are not provided here. synchronization mechanisms that are not provided here.
An endpoint MAY bundle PATH_CHALLENGE and PATH_RESPONSE frames that An endpoint MAY bundle PATH_CHALLENGE and PATH_RESPONSE frames that
are used for path validation with other frames. For instance, an are used for path validation with other frames. In particular, an
endpoint may pad a packet carrying a PATH_CHALLENGE for PMTU endpoint may pad a packet carrying a PATH_CHALLENGE for PMTU
discovery, or an endpoint may bundle a PATH_RESPONSE with its own discovery, or an endpoint may bundle a PATH_RESPONSE with its own
PATH_CHALLENGE. PATH_CHALLENGE.
When probing a new path, an endpoint might want to ensure that its When probing a new path, an endpoint might want to ensure that its
peer has an unused connection ID available for responses. The peer has an unused connection ID available for responses. The
endpoint can send NEW_CONNECTION_ID and PATH_CHALLENGE frames in the endpoint can send NEW_CONNECTION_ID and PATH_CHALLENGE frames in the
same packet. This ensures that an unused connection ID will be same packet. This ensures that an unused connection ID will be
available to the peer when sending a response. available to the peer when sending a response.
6.10.1. Initiation 8.3. Initiating Path Validation
To initiate path validation, an endpoint sends a PATH_CHALLENGE frame To initiate path validation, an endpoint sends a PATH_CHALLENGE frame
containing a random payload on the path to be validated. containing a random payload on the path to be validated.
An endpoint MAY send additional PATH_CHALLENGE frames to handle An endpoint MAY send multiple PATH_CHALLENGE frames to guard against
packet loss. An endpoint SHOULD NOT send a PATH_CHALLENGE more packet loss. An endpoint SHOULD NOT send a PATH_CHALLENGE more
frequently than it would an Initial packet, ensuring that connection frequently than it would an Initial packet, ensuring that connection
migration is no more load on a new path than establishing a new migration is no more load on a new path than establishing a new
connection. connection.
The endpoint MUST use fresh random data in every PATH_CHALLENGE frame The endpoint MUST use fresh random data in every PATH_CHALLENGE frame
so that it can associate the peer's response with the causative so that it can associate the peer's response with the causative
PATH_CHALLENGE. PATH_CHALLENGE.
6.10.2. Response 8.4. Path Validation Responses
On receiving a PATH_CHALLENGE frame, an endpoint MUST respond On receiving a PATH_CHALLENGE frame, an endpoint MUST respond
immediately by echoing the data contained in the PATH_CHALLENGE frame immediately by echoing the data contained in the PATH_CHALLENGE frame
in a PATH_RESPONSE frame, with the following stipulation. Since a in a PATH_RESPONSE frame. However, because a PATH_CHALLENGE might be
PATH_CHALLENGE might be sent from a spoofed address, an endpoint MAY sent from a spoofed address, an endpoint MUST limit the rate at which
limit the rate at which it sends PATH_RESPONSE frames and MAY it sends PATH_RESPONSE frames and MAY silently discard PATH_CHALLENGE
silently discard PATH_CHALLENGE frames that would cause it to respond frames that would cause it to respond at a higher rate.
at a higher rate.
To ensure that packets can be both sent to and received from the To ensure that packets can be both sent to and received from the
peer, the PATH_RESPONSE MUST be sent on the same path as the peer, the PATH_RESPONSE MUST be sent on the same path as the
triggering PATH_CHALLENGE: from the same local address on which the triggering PATH_CHALLENGE. That is, from the same local address on
PATH_CHALLENGE was received, to the same remote address from which which the PATH_CHALLENGE was received, to the same remote address
the PATH_CHALLENGE was received. from which the PATH_CHALLENGE was received.
6.10.3. Completion 8.5. Successful Path Validation
A new address is considered valid when a PATH_RESPONSE frame is A new address is considered valid when a PATH_RESPONSE frame is
received containing data that was sent in a previous PATH_CHALLENGE. received containing data that was sent in a previous PATH_CHALLENGE.
Receipt of an acknowledgment for a packet containing a PATH_CHALLENGE Receipt of an acknowledgment for a packet containing a PATH_CHALLENGE
frame is not adequate validation, since the acknowledgment can be frame is not adequate validation, since the acknowledgment can be
spoofed by a malicious peer. spoofed by a malicious peer.
For path validation to be successful, a PATH_RESPONSE frame MUST be For path validation to be successful, a PATH_RESPONSE frame MUST be
received from the same remote address to which the corresponding received from the same remote address to which the corresponding
PATH_CHALLENGE was sent. If a PATH_RESPONSE frame is received from a PATH_CHALLENGE was sent. If a PATH_RESPONSE frame is received from a
skipping to change at page 53, line 5 skipping to change at page 43, line 23
Additionally, the PATH_RESPONSE frame MUST be received on the same Additionally, the PATH_RESPONSE frame MUST be received on the same
local address from which the corresponding PATH_CHALLENGE was sent. local address from which the corresponding PATH_CHALLENGE was sent.
If a PATH_RESPONSE frame is received on a different local address If a PATH_RESPONSE frame is received on a different local address
than the one from which the PATH_CHALLENGE was sent, path validation than the one from which the PATH_CHALLENGE was sent, path validation
is considered to have failed, even if the data matches that sent in is considered to have failed, even if the data matches that sent in
the PATH_CHALLENGE. Thus, the endpoint considers the path to be the PATH_CHALLENGE. Thus, the endpoint considers the path to be
valid when a PATH_RESPONSE frame is received on the same path with valid when a PATH_RESPONSE frame is received on the same path with
the same payload as the PATH_CHALLENGE frame. the same payload as the PATH_CHALLENGE frame.
6.10.4. Abandonment 8.6. Failed Path Validation
An endpoint SHOULD abandon path validation after sending some number Path validation only fails when the endpoint attempting to validate
of PATH_CHALLENGE frames or after some time has passed. When setting the path abandons its attempt to validate the path.
this timer, implementations are cautioned that the new path could
have a longer round-trip time than the original. Endpoints SHOULD abandon path validation based on a timer. When
setting this timer, implementations are cautioned that the new path
could have a longer round-trip time than the original.
Note that the endpoint might receive packets containing other frames Note that the endpoint might receive packets containing other frames
on the new path, but a PATH_RESPONSE frame with appropriate data is on the new path, but a PATH_RESPONSE frame with appropriate data is
required for path validation to succeed. required for path validation to succeed.
If path validation fails, the path is deemed unusable. This does not When an endpoint abandons path validation, it determines that the
necessarily imply a failure of the connection - endpoints can path is unusable. This does not necessarily imply a failure of the
continue sending packets over other paths as appropriate. If no connection - endpoints can continue sending packets over other paths
paths are available, an endpoint can wait for a new path to become as appropriate. If no paths are available, an endpoint can wait for
available or close the connection. a new path to become available or close the connection.
A path validation might be abandoned for other reasons besides A path validation might be abandoned for other reasons besides
failure. Primarily, this happens if a connection migration to a new failure. Primarily, this happens if a connection migration to a new
path is initiated while a path validation on the old path is in path is initiated while a path validation on the old path is in
progress. progress.
6.11. Connection Migration 9. Connection Migration
QUIC allows connections to survive changes to endpoint addresses The use of a connection ID allows connections to survive changes to
(that is, IP address and/or port), such as those caused by an endpoint addresses (that is, IP address and/or port), such as those
endpoint migrating to a new network. This section describes the caused by an endpoint migrating to a new network. This section
process by which an endpoint migrates to a new address. describes the process by which an endpoint migrates to a new address.
An endpoint MUST NOT initiate connection migration before the An endpoint MUST NOT initiate connection migration before the
handshake is finished and the endpoint has 1-RTT keys. The design of handshake is finished and the endpoint has 1-RTT keys. The design of
QUIC relies on endpoints retaining a stable address for the duration QUIC relies on endpoints retaining a stable address for the duration
of the handshake. of the handshake.
An endpoint also MUST NOT initiate connection migration if the peer An endpoint also MUST NOT initiate connection migration if the peer
sent the "disable_migration" transport parameter during the sent the "disable_migration" transport parameter during the
handshake. An endpoint which has sent this transport parameter, but handshake. An endpoint which has sent this transport parameter, but
detects that a peer has nonetheless migrated to a different network detects that a peer has nonetheless migrated to a different network
MAY treat this as a connection error of type INVALID_MIGRATION. MAY treat this as a connection error of type INVALID_MIGRATION.
Not all changes of peer address are intentional migrations. The peer Not all changes of peer address are intentional migrations. The peer
could experience NAT rebinding: a change of address due to a could experience NAT rebinding: a change of address due to a
middlebox, usually a NAT, allocating a new outgoing port or even a middlebox, usually a NAT, allocating a new outgoing port or even a
new outgoing IP address for a flow. Endpoints SHOULD perform path new outgoing IP address for a flow. NAT rebinding is not connection
validation (Section 6.10) if a NAT rebinding does not cause the migration as defined in this section, though an endpoint SHOULD
connection to fail. perform path validation (Section 8.2) if it detects a change in the
IP address of its peer.
This document limits migration of connections to new client This document limits migration of connections to new client
addresses, except as described in Section 6.12. Clients are addresses, except as described in Section 9.6. Clients are
responsible for initiating all migrations. Servers do not send non- responsible for initiating all migrations. Servers do not send non-
probing packets (see Section 6.11.1) toward a client address until probing packets (see Section 9.1) toward a client address until they
they see a non-probing packet from that address. If a client see a non-probing packet from that address. If a client receives
receives packets from an unknown server address, the client MAY packets from an unknown server address, the client MAY discard these
discard these packets. packets.
6.11.1. Probing a New Path 9.1. Probing a New Path
An endpoint MAY probe for peer reachability from a new local address An endpoint MAY probe for peer reachability from a new local address
using path validation Section 6.10 prior to migrating the connection using path validation Section 8.2 prior to migrating the connection
to the new local address. Failure of path validation simply means to the new local address. Failure of path validation simply means
that the new path is not usable for this connection. Failure to that the new path is not usable for this connection. Failure to
validate a path does not cause the connection to end unless there are validate a path does not cause the connection to end unless there are
no valid alternative paths available. no valid alternative paths available.
An endpoint uses a new connection ID for probes sent from a new local An endpoint uses a new connection ID for probes sent from a new local
address, see Section 6.11.5 for further discussion. An endpoint that address, see Section 9.5 for further discussion. An endpoint that
uses a new local address needs to ensure that at least one new uses a new local address needs to ensure that at least one new
connection ID is available at the peer. That can be achieved by connection ID is available at the peer. That can be achieved by
including a NEW_CONNECTION_ID frame in the probe. including a NEW_CONNECTION_ID frame in the probe.
Receiving a PATH_CHALLENGE frame from a peer indicates that the peer Receiving a PATH_CHALLENGE frame from a peer indicates that the peer
is probing for reachability on a path. An endpoint sends a is probing for reachability on a path. An endpoint sends a
PATH_RESPONSE in response as per Section 6.10. PATH_RESPONSE in response as per Section 8.2.
PATH_CHALLENGE, PATH_RESPONSE, NEW_CONNECTION_ID, and PADDING frames PATH_CHALLENGE, PATH_RESPONSE, NEW_CONNECTION_ID, and PADDING frames
are "probing frames", and all other frames are "non-probing frames". are "probing frames", and all other frames are "non-probing frames".
A packet containing only probing frames is a "probing packet", and a A packet containing only probing frames is a "probing packet", and a
packet containing any other frame is a "non-probing packet". packet containing any other frame is a "non-probing packet".
6.11.2. Initiating Connection Migration 9.2. Initiating Connection Migration
An endpoint can migrate a connection to a new local address by An endpoint can migrate a connection to a new local address by
sending packets containing frames other than probing frames from that sending packets containing non-probing frames from that address.
address.
Each endpoint validates its peer's address during connection Each endpoint validates its peer's address during connection
establishment. Therefore, a migrating endpoint can send to its peer establishment. Therefore, a migrating endpoint can send to its peer
knowing that the peer is willing to receive at the peer's current knowing that the peer is willing to receive at the peer's current
address. Thus an endpoint can migrate to a new local address without address. Thus an endpoint can migrate to a new local address without
first validating the peer's address. first validating the peer's address.
When migrating, the new path might not support the endpoint's current When migrating, the new path might not support the endpoint's current
sending rate. Therefore, the endpoint resets its congestion sending rate. Therefore, the endpoint resets its congestion
controller, as described in Section 6.11.4. controller, as described in Section 9.4.
The new path might not have the same ECN capability. Therefore, the The new path might not have the same ECN capability. Therefore, the
endpoint verifies ECN capability as described in Section 6.8. endpoint verifies ECN capability as described in Section 13.3.
Receiving acknowledgments for data sent on the new path serves as Receiving acknowledgments for data sent on the new path serves as
proof of the peer's reachability from the new address. Note that proof of the peer's reachability from the new address. Note that
since acknowledgments may be received on any path, return since acknowledgments may be received on any path, return
reachability on the new path is not established. To establish return reachability on the new path is not established. To establish return
reachability on the new path, an endpoint MAY concurrently initiate reachability on the new path, an endpoint MAY concurrently initiate
path validation Section 6.10 on the new path. path validation Section 8.2 on the new path.
6.11.3. Responding to Connection Migration 9.3. Responding to Connection Migration
Receiving a packet from a new peer address containing a non-probing Receiving a packet from a new peer address containing a non-probing
frame indicates that the peer has migrated to that address. frame indicates that the peer has migrated to that address.
In response to such a packet, an endpoint MUST start sending In response to such a packet, an endpoint MUST start sending
subsequent packets to the new peer address and MUST initiate path subsequent packets to the new peer address and MUST initiate path
validation (Section 6.10) to verify the peer's ownership of the validation (Section 8.2) to verify the peer's ownership of the
unvalidated address. unvalidated address.
An endpoint MAY send data to an unvalidated peer address, but it MUST An endpoint MAY send data to an unvalidated peer address, but it MUST
protect against potential attacks as described in Section 6.11.3.1 protect against potential attacks as described in Section 9.3.1 and
and Section 6.11.3.2. An endpoint MAY skip validation of a peer Section 9.3.2. An endpoint MAY skip validation of a peer address if
address if that address has been seen recently. that address has been seen recently.
An endpoint only changes the address that it sends packets to in An endpoint only changes the address that it sends packets to in
response to the highest-numbered non-probing packet. This ensures response to the highest-numbered non-probing packet. This ensures
that an endpoint does not send packets to an old peer address in the that an endpoint does not send packets to an old peer address in the
case that it receives reordered packets. case that it receives reordered packets.
After changing the address to which it sends non-probing packets, an After changing the address to which it sends non-probing packets, an
endpoint could abandon any path validation for other addresses. endpoint could abandon any path validation for other addresses.
Receiving a packet from a new peer address might be the result of a Receiving a packet from a new peer address might be the result of a
NAT rebinding at the peer. NAT rebinding at the peer.
After verifying a new client address, the server SHOULD send new After verifying a new client address, the server SHOULD send new
address validation tokens (Section 6.9) to the client. address validation tokens (Section 8) to the client.
6.11.3.1. Handling Address Spoofing by a Peer 9.3.1. Handling Address Spoofing by a Peer
It is possible that a peer is spoofing its source address to cause an It is possible that a peer is spoofing its source address to cause an
endpoint to send excessive amounts of data to an unwilling host. If endpoint to send excessive amounts of data to an unwilling host. If
the endpoint sends significantly more data than the spoofing peer, the endpoint sends significantly more data than the spoofing peer,
connection migration might be used to amplify the volume of data that connection migration might be used to amplify the volume of data that
an attacker can generate toward a victim. an attacker can generate toward a victim.
As described in Section 6.11.3, an endpoint is required to validate a As described in Section 9.3, an endpoint is required to validate a
peer's new address to confirm the peer's possession of the new peer's new address to confirm the peer's possession of the new
address. Until a peer's address is deemed valid, an endpoint MUST address. Until a peer's address is deemed valid, an endpoint MUST
limit the rate at which it sends data to this address. The endpoint limit the rate at which it sends data to this address. The endpoint
MUST NOT send more than a minimum congestion window's worth of data MUST NOT send more than a minimum congestion window's worth of data
per estimated round-trip time (kMinimumWindow, as defined in per estimated round-trip time (kMinimumWindow, as defined in
[QUIC-RECOVERY]). In the absence of this limit, an endpoint risks [QUIC-RECOVERY]). In the absence of this limit, an endpoint risks
being used for a denial of service attack against an unsuspecting being used for a denial of service attack against an unsuspecting
victim. Note that since the endpoint will not have any round-trip victim. Note that since the endpoint will not have any round-trip
time measurements to this address, the estimate SHOULD be the default time measurements to this address, the estimate SHOULD be the default
initial value (see [QUIC-RECOVERY]). initial value (see [QUIC-RECOVERY]).
If an endpoint skips validation of a peer address as described in If an endpoint skips validation of a peer address as described in
Section 6.11.3, it does not need to limit its sending rate. Section 9.3, it does not need to limit its sending rate.
6.11.3.2. Handling Address Spoofing by an On-path Attacker 9.3.2. Handling Address Spoofing by an On-path Attacker
An on-path attacker could cause a spurious connection migration by An on-path attacker could cause a spurious connection migration by
copying and forwarding a packet with a spoofed address such that it copying and forwarding a packet with a spoofed address such that it
arrives before the original packet. The packet with the spoofed arrives before the original packet. The packet with the spoofed
address will be seen to come from a migrating connection, and the address will be seen to come from a migrating connection, and the
original packet will be seen as a duplicate and dropped. After a original packet will be seen as a duplicate and dropped. After a
spurious migration, validation of the source address will fail spurious migration, validation of the source address will fail
because the entity at the source address does not have the necessary because the entity at the source address does not have the necessary
cryptographic keys to read or respond to the PATH_CHALLENGE frame cryptographic keys to read or respond to the PATH_CHALLENGE frame
that is sent to it even if it wanted to. that is sent to it even if it wanted to.
skipping to change at page 56, line 47 skipping to change at page 47, line 20
MUST close the connection silently by discarding all connection MUST close the connection silently by discarding all connection
state. This results in new packets on the connection being handled state. This results in new packets on the connection being handled
generically. For instance, an endpoint MAY send a stateless reset in generically. For instance, an endpoint MAY send a stateless reset in
response to any further incoming packets. response to any further incoming packets.
Note that receipt of packets with higher packet numbers from the Note that receipt of packets with higher packet numbers from the
legitimate peer address will trigger another connection migration. legitimate peer address will trigger another connection migration.
This will cause the validation of the address of the spurious This will cause the validation of the address of the spurious
migration to be abandoned. migration to be abandoned.
6.11.4. Loss Detection and Congestion Control 9.4. Loss Detection and Congestion Control
The capacity available on the new path might not be the same as the The capacity available on the new path might not be the same as the
old path. Packets sent on the old path SHOULD NOT contribute to old path. Packets sent on the old path SHOULD NOT contribute to
congestion control or RTT estimation for the new path. congestion control or RTT estimation for the new path.
On confirming a peer's ownership of its new address, an endpoint On confirming a peer's ownership of its new address, an endpoint
SHOULD immediately reset the congestion controller and round-trip SHOULD immediately reset the congestion controller and round-trip
time estimator for the new path. time estimator for the new path.
An endpoint MUST NOT return to the send rate used for the previous An endpoint MUST NOT return to the send rate used for the previous
skipping to change at page 57, line 36 skipping to change at page 48, line 8
single congestion control context and a single loss recovery context single congestion control context and a single loss recovery context
(as described in [QUIC-RECOVERY]) may be adequate. A sender can make (as described in [QUIC-RECOVERY]) may be adequate. A sender can make
exceptions for probe packets so that their loss detection is exceptions for probe packets so that their loss detection is
independent and does not unduly cause the congestion controller to independent and does not unduly cause the congestion controller to
reduce its sending rate. An endpoint might set a separate timer when reduce its sending rate. An endpoint might set a separate timer when
a PATH_CHALLENGE is sent, which is cancelled when the corresponding a PATH_CHALLENGE is sent, which is cancelled when the corresponding
PATH_RESPONSE is received. If the timer fires before the PATH_RESPONSE is received. If the timer fires before the
PATH_RESPONSE is received, the endpoint might send a new PATH_RESPONSE is received, the endpoint might send a new
PATH_CHALLENGE, and restart the timer for a longer period of time. PATH_CHALLENGE, and restart the timer for a longer period of time.
6.11.5. Privacy Implications of Connection Migration 9.5. Privacy Implications of Connection Migration
Using a stable connection ID on multiple network paths allows a Using a stable connection ID on multiple network paths allows a
passive observer to correlate activity between those paths. An passive observer to correlate activity between those paths. An
endpoint that moves between networks might not wish to have their endpoint that moves between networks might not wish to have their
activity correlated by any entity other than their peer, so different activity correlated by any entity other than their peer, so different
connection IDs are used when sending from different local addresses, connection IDs are used when sending from different local addresses,
as discussed in Section 6.1. For this to be effective endpoints need as discussed in Section 5.1. For this to be effective endpoints need
to ensure that connections IDs they provide cannot be linked by any to ensure that connections IDs they provide cannot be linked by any
other entity. other entity.
This eliminates the use of the connection ID for linking activity This eliminates the use of the connection ID for linking activity
from the same connection on different networks. Protection of packet from the same connection on different networks. Protection of packet
numbers ensures that packet numbers cannot be used to correlate numbers ensures that packet numbers cannot be used to correlate
activity. This does not prevent other properties of packets, such as activity. This does not prevent other properties of packets, such as
timing and size, from being used to correlate activity. timing and size, from being used to correlate activity.
Clients MAY move to a new connection ID at any time based on Clients MAY move to a new connection ID at any time based on
skipping to change at page 58, line 17 skipping to change at page 48, line 37
network inactivity NAT rebinding might occur when the client begins network inactivity NAT rebinding might occur when the client begins
sending data again. sending data again.
A client might wish to reduce linkability by employing a new A client might wish to reduce linkability by employing a new
connection ID and source UDP port when sending traffic after a period connection ID and source UDP port when sending traffic after a period
of inactivity. Changing the UDP port from which it sends packets at of inactivity. Changing the UDP port from which it sends packets at
the same time might cause the packet to appear as a connection the same time might cause the packet to appear as a connection
migration. This ensures that the mechanisms that support migration migration. This ensures that the mechanisms that support migration
are exercised even for clients that don't experience NAT rebindings are exercised even for clients that don't experience NAT rebindings
or genuine migrations. Changing port number can cause a peer to or genuine migrations. Changing port number can cause a peer to
reset its congestion state (see Section 6.11.4), so the port SHOULD reset its congestion state (see Section 9.4), so the port SHOULD only
only be changed infrequently. be changed infrequently.
Endpoints that use connection IDs with length greater than zero could Endpoints that use connection IDs with length greater than zero could
have their activity correlated if their peers keep using the same have their activity correlated if their peers keep using the same
destination connection ID after migration. Endpoints that receive destination connection ID after migration. Endpoints that receive
packets with a previously unused Destination Connection ID SHOULD packets with a previously unused Destination Connection ID SHOULD
change to sending packets with a connection ID that has not been used change to sending packets with a connection ID that has not been used
on any other network path. The goal here is to ensure that packets on any other network path. The goal here is to ensure that packets
sent on different paths cannot be correlated. To fulfill this sent on different paths cannot be correlated. To fulfill this
privacy requirement, endpoints that initiate migration and use privacy requirement, endpoints that initiate migration and use
connection IDs with length greater than zero SHOULD provide their connection IDs with length greater than zero SHOULD provide their
peers with new connection IDs before migration. peers with new connection IDs before migration.
Caution: If both endpoints change connection ID in response to Caution: If both endpoints change connection ID in response to
seeing a change in connection ID from their peer, then this can seeing a change in connection ID from their peer, then this can
trigger an infinite sequence of changes. trigger an infinite sequence of changes.
6.12. Server's Preferred Address 9.6. Server's Preferred Address
QUIC allows servers to accept connections on one IP address and QUIC allows servers to accept connections on one IP address and
attempt to transfer these connections to a more preferred address attempt to transfer these connections to a more preferred address
shortly after the handshake. This is particularly useful when shortly after the handshake. This is particularly useful when
clients initially connect to an address shared by multiple servers clients initially connect to an address shared by multiple servers
but would prefer to use a unicast address to ensure connection but would prefer to use a unicast address to ensure connection
stability. This section describes the protocol for migrating a stability. This section describes the protocol for migrating a
connection to a preferred server address. connection to a preferred server address.
Migrating a connection to a new server address mid-connection is left Migrating a connection to a new server address mid-connection is left
for future work. If a client receives packets from a new server for future work. If a client receives packets from a new server
address not indicated by the preferred_address transport parameter, address not indicated by the preferred_address transport parameter,
the client SHOULD discard these packets. the client SHOULD discard these packets.
6.12.1. Communicating A Preferred Address 9.6.1. Communicating A Preferred Address
A server conveys a preferred address by including the A server conveys a preferred address by including the
preferred_address transport parameter in the TLS handshake. preferred_address transport parameter in the TLS handshake.
Once the handshake is finished, the client SHOULD initiate path Once the handshake is finished, the client SHOULD initiate path
validation (see Section 6.10) of the server's preferred address using validation (see Section 8.2) of the server's preferred address using
the connection ID provided in the preferred_address transport the connection ID provided in the preferred_address transport
parameter. parameter.
If path validation succeeds, the client SHOULD immediately begin If path validation succeeds, the client SHOULD immediately begin
sending all future packets to the new server address using the new sending all future packets to the new server address using the new
connection ID and discontinue use of the old server address. If path connection ID and discontinue use of the old server address. If path
validation fails, the client MUST continue sending all future packets validation fails, the client MUST continue sending all future packets
to the server's original IP address. to the server's original IP address.
6.12.2. Responding to Connection Migration 9.6.2. Responding to Connection Migration
A server might receive a packet addressed to its preferred IP address A server might receive a packet addressed to its preferred IP address
at any time after the handshake is completed. If this packet at any time after it accepts a connection. If this packet contains a
contains a PATH_CHALLENGE frame, the server sends a PATH_RESPONSE PATH_CHALLENGE frame, the server sends a PATH_RESPONSE frame as per
frame as per Section 6.10, but the server MUST continue sending all Section 8.2. The server MAY send other non-probing frames from its
other packets from its original IP address. preferred address, but MUST continue sending all probing packets from
its original IP address.
The server SHOULD also initiate path validation of the client using The server SHOULD also initiate path validation of the client using
its preferred address and the address from which it received the its preferred address and the address from which it received the
client probe. This helps to guard against spurious migration client probe. This helps to guard against spurious migration
initiated by an attacker. initiated by an attacker.
Once the server has completed its path validation and has received a Once the server has completed its path validation and has received a
non-probing packet with a new largest packet number on its preferred non-probing packet with a new largest packet number on its preferred
address, the server begins sending to the client exclusively from its address, the server begins sending non-probing packets to the client
preferred IP address. It SHOULD drop packets for this connection exclusively from its preferred IP address. It SHOULD drop packets
received on the old IP address, but MAY continue to process delayed for this connection received on the old IP address, but MAY continue
packets. to process delayed packets.
6.12.3. Interaction of Client Migration and Preferred Address 9.6.3. Interaction of Client Migration and Preferred Address
A client might need to perform a connection migration before it has A client might need to perform a connection migration before it has
migrated to the server's preferred address. In this case, the client migrated to the server's preferred address. In this case, the client
SHOULD perform path validation to both the original and preferred SHOULD perform path validation to both the original and preferred
server address from the client's new address concurrently. server address from the client's new address concurrently.
If path validation of the server's preferred address succeeds, the If path validation of the server's preferred address succeeds, the
client MUST abandon validation of the original address and migrate to client MUST abandon validation of the original address and migrate to
using the server's preferred address. If path validation of the using the server's preferred address. If path validation of the
server's preferred address fails, but validation of the server's server's preferred address fails but validation of the server's
original address succeeds, the client MAY migrate to using the original address succeeds, the client MAY migrate to its new address
original address from the client's new address. and continue sending to the server's original address.
If the connection to the server's preferred address is not from the If the connection to the server's preferred address is not from the
same client address, the server MUST protect against potential same client address, the server MUST protect against potential
attacks as described in Section 6.11.3.1 and Section 6.11.3.2. In attacks as described in Section 9.3.1 and Section 9.3.2. In addition
addition to intentional simultaneous migration, this might also occur to intentional simultaneous migration, this might also occur because
because the client's access network used a different NAT binding for the client's access network used a different NAT binding for the
the server's preferred address. server's preferred address.
Servers SHOULD initiate path validation to the client's new address Servers SHOULD initiate path validation to the client's new address
upon receiving a probe packet from a different address. Servers MUST upon receiving a probe packet from a different address. Servers MUST
NOT send more than a minimum congestion window's worth of non-probing NOT send more than a minimum congestion window's worth of non-probing
packets to the new address before path validation is complete. packets to the new address before path validation is complete.
6.13. Connection Termination 10. Connection Termination
Connections should remain open until they become idle for a pre- Connections should remain open until they become idle for a pre-
negotiated period of time. A QUIC connection, once established, can negotiated period of time. A QUIC connection, once established, can
be terminated in one of three ways: be terminated in one of three ways:
o idle timeout (Section 6.13.2) o idle timeout (Section 10.2)
o immediate close (Section 6.13.3) o immediate close (Section 10.3)
o stateless reset (Section 6.13.4) o stateless reset (Section 10.4)
6.13.1. Closing and Draining Connection States 10.1. Closing and Draining Connection States
The closing and draining connection states exist to ensure that The closing and draining connection states exist to ensure that
connections close cleanly and that delayed or reordered packets are connections close cleanly and that delayed or reordered packets are
properly discarded. These states SHOULD persist for three times the properly discarded. These states SHOULD persist for three times the
current Retransmission Timeout (RTO) interval as defined in current Retransmission Timeout (RTO) interval as defined in
[QUIC-RECOVERY]. [QUIC-RECOVERY].
An endpoint enters a closing period after initiating an immediate An endpoint enters a closing period after initiating an immediate
close (Section 6.13.3). While closing, an endpoint MUST NOT send close (Section 10.3). While closing, an endpoint MUST NOT send
packets unless they contain a CONNECTION_CLOSE or APPLICATION_CLOSE packets unless they contain a CONNECTION_CLOSE or APPLICATION_CLOSE
frame (see Section 6.13.3 for details). frame (see Section 10.3 for details). An endpoint retains only
enough information to generate a packet containing a closing frame
In the closing state, only a packet containing a closing frame can be and to identify packets as belonging to the connection. The
sent. An endpoint retains only enough information to generate a connection ID and QUIC version is sufficient information to identify
packet containing a closing frame and to identify packets as packets for a closing connection; an endpoint can discard all other
belonging to the connection. The connection ID and QUIC version is connection state. An endpoint MAY retain packet protection keys for
sufficient information to identify packets for a closing connection; incoming packets to allow it to read and process a closing frame.
an endpoint can discard all other connection state. An endpoint MAY
retain packet protection keys for incoming packets to allow it to
read and process a closing frame.
The draining state is entered once an endpoint receives a signal that The draining state is entered once an endpoint receives a signal that
its peer is closing or draining. While otherwise identical to the its peer is closing or draining. While otherwise identical to the
closing state, an endpoint in the draining state MUST NOT send any closing state, an endpoint in the draining state MUST NOT send any
packets. Retaining packet protection keys is unnecessary once a packets. Retaining packet protection keys is unnecessary once a
connection is in the draining state. connection is in the draining state.
An endpoint MAY transition from the closing period to the draining An endpoint MAY transition from the closing period to the draining
period if it can confirm that its peer is also closing or draining. period if it can confirm that its peer is also closing or draining.
Receiving a closing frame is sufficient confirmation, as is receiving Receiving a closing frame is sufficient confirmation, as is receiving
skipping to change at page 61, line 33 skipping to change at page 51, line 52
an abbreviated draining period which can allow for faster resource an abbreviated draining period which can allow for faster resource
recovery. Servers that retain an open socket for accepting new recovery. Servers that retain an open socket for accepting new
connections SHOULD NOT exit the closing or draining period early. connections SHOULD NOT exit the closing or draining period early.
Once the closing or draining period has ended, an endpoint SHOULD Once the closing or draining period has ended, an endpoint SHOULD
discard all connection state. This results in new packets on the discard all connection state. This results in new packets on the
connection being handled generically. For instance, an endpoint MAY connection being handled generically. For instance, an endpoint MAY
send a stateless reset in response to any further incoming packets. send a stateless reset in response to any further incoming packets.
The draining and closing periods do not apply when a stateless reset The draining and closing periods do not apply when a stateless reset
(Section 6.13.4) is sent. (Section 10.4) is sent.
An endpoint is not expected to handle key updates when it is closing An endpoint is not expected to handle key updates when it is closing
or draining. A key update might prevent the endpoint from moving or draining. A key update might prevent the endpoint from moving
from the closing state to draining, but it otherwise has no impact. from the closing state to draining, but it otherwise has no impact.
An endpoint could receive packets from a new source address, An endpoint could receive packets from a new source address,
indicating a client connection migration (Section 6.11), while in the indicating a client connection migration (Section 9), while in the
closing period. An endpoint in the closing state MUST strictly limit closing period. An endpoint in the closing state MUST strictly limit
the number of packets it sends to this new address until the address the number of packets it sends to this new address until the address
is validated (see Section 6.10). A server in the closing state MAY is validated (see Section 8.2). A server in the closing state MAY
instead choose to discard packets received from a new source address. instead choose to discard packets received from a new source address.
6.13.2. Idle Timeout 10.2. Idle Timeout
If the idle timeout is enabled, a connection that remains idle for If the idle timeout is enabled, a connection that remains idle for
longer than the advertised idle timeout (see Section 6.6.1) is longer than the advertised idle timeout (see Section 18.1) is closed.
closed. A connection enters the draining state when the idle timeout A connection enters the draining state when the idle timeout expires.
expires.
Each endpoint advertises their own idle timeout to their peer. The Each endpoint advertises its own idle timeout to its peer. The idle
idle timeout starts from the last packet received. In order to timeout starts from the last packet received. In order to ensure
ensure that initiating new activity postpones an idle timeout, an that initiating new activity postpones an idle timeout, an endpoint
endpoint restarts this timer when sending a packet. An endpoint does restarts this timer when sending a packet. An endpoint does not
not postpone the idle timeout if another packet has been sent postpone the idle timeout if another packet has been sent containing
containing frames other than ACK or PADDING, and that other packet frames other than ACK or PADDING, and that other packet has not been
has not been acknowledged or declared lost. Packets that contain acknowledged or declared lost. Packets that contain only ACK or
only ACK or PADDING frames are not acknowledged until an endpoint has PADDING frames are not acknowledged until an endpoint has other
other frames to send, so they could prevent the timeout from being frames to send, so they could prevent the timeout from being
refreshed. refreshed.
The value for an idle timeout can be asymmetric. The value The value for an idle timeout can be asymmetric. The value
advertised by an endpoint is only used to determine whether the advertised by an endpoint is only used to determine whether the
connection is live at that endpoint. An endpoint that sends packets connection is live at that endpoint. An endpoint that sends packets
near the end of the idle timeout period of a peer risks having those near the end of the idle timeout period of a peer risks having those
packets discarded if its peer enters the draining state before the packets discarded if its peer enters the draining state before the
packets arrive. If a peer could timeout within an RTO (see packets arrive. If a peer could timeout within an RTO (see
Section 4.3.3 of [QUIC-RECOVERY]), it is advisable to test for Section 4.3.3 of [QUIC-RECOVERY]), it is advisable to test for
liveness before sending any data that cannot be retried safely. liveness before sending any data that cannot be retried safely.
6.13.3. Immediate Close 10.3. Immediate Close
An endpoint sends a closing frame (CONNECTION_CLOSE or An endpoint sends a closing frame (CONNECTION_CLOSE or
APPLICATION_CLOSE) to terminate the connection immediately. Any APPLICATION_CLOSE) to terminate the connection immediately. Any
closing frame causes all streams to immediately become closed; open closing frame causes all streams to immediately become closed; open
streams can be assumed to be implicitly reset. streams can be assumed to be implicitly reset.
After sending a closing frame, endpoints immediately enter the After sending a closing frame, endpoints immediately enter the
closing state. During the closing period, an endpoint that sends a closing state. During the closing period, an endpoint that sends a
closing frame SHOULD respond to any packet that it receives with closing frame SHOULD respond to any packet that it receives with
another packet containing a closing frame. To minimize the state another packet containing a closing frame. To minimize the state
skipping to change at page 63, line 18 skipping to change at page 53, line 35
An immediate close can be used after an application protocol has An immediate close can be used after an application protocol has
arranged to close a connection. This might be after the application arranged to close a connection. This might be after the application
protocols negotiates a graceful shutdown. The application protocol protocols negotiates a graceful shutdown. The application protocol
exchanges whatever messages that are needed to cause both endpoints exchanges whatever messages that are needed to cause both endpoints
to agree to close the connection, after which the application to agree to close the connection, after which the application
requests that the connection be closed. The application protocol can requests that the connection be closed. The application protocol can
use an APPLICATION_CLOSE message with an appropriate error code to use an APPLICATION_CLOSE message with an appropriate error code to
signal closure. signal closure.
6.13.4. Stateless Reset If the connection has been successfully established, endpoints MUST
send any closing frames in a 1-RTT packet. Prior to connection
establishment a peer might not have 1-RTT keys, so endpoints SHOULD
send closing frames in a Handshake packet. If the endpoint does not
have Handshake keys, or it is not certain that the peer has Handshake
keys, it MAY send closing frames in an Initial packet. If multiple
packets are sent, they can be coalesced (see Section 12.2) to
facilitate retransmission.
10.4. Stateless Reset
A stateless reset is provided as an option of last resort for an A stateless reset is provided as an option of last resort for an
endpoint that does not have access to the state of a connection. A endpoint that does not have access to the state of a connection. A
crash or outage might result in peers continuing to send data to an crash or outage might result in peers continuing to send data to an
endpoint that is unable to properly continue the connection. An endpoint that is unable to properly continue the connection. An
endpoint that wishes to communicate a fatal connection error MUST use endpoint that wishes to communicate a fatal connection error MUST use
a closing frame if it has sufficient state to do so. a closing frame if it has sufficient state to do so.
To support this process, a token is sent by endpoints. The token is To support this process, a token is sent by endpoints. The token is
carried in the NEW_CONNECTION_ID frame sent by either peer, and carried in the NEW_CONNECTION_ID frame sent by either peer, and
servers can specify the stateless_reset_token transport parameter servers can specify the stateless_reset_token transport parameter
during the handshake (clients cannot because their transport during the handshake (clients cannot because their transport
parameters don't have confidentiality protection). This value is parameters don't have confidentiality protection). This value is
protected by encryption, so only client and server know this value. protected by encryption, so only client and server know this value.
Tokens sent via NEW_CONNECTION_ID frames are invalidated when their Tokens sent via NEW_CONNECTION_ID frames are invalidated when their
associated connection ID is retired via a RETIRE_CONNECTION_ID frame associated connection ID is retired via a RETIRE_CONNECTION_ID frame
(Section 7.14). (Section 19.13).
An endpoint that receives packets that it cannot process sends a An endpoint that receives packets that it cannot process sends a
packet in the following layout: packet in the following layout:
0 1 2 3 0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+
|0|K|1|1|0|0|0|0| |0|K|1|1|0|0|0|0|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Random Octets (160..) ... | Random Octets (160..) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| | | |
+ + + +
| | | |
+ Stateless Reset Token (128) + + Stateless Reset Token (128) +
| | | |
+ + + +
| | | |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 11: Stateless Reset Packet Figure 7: Stateless Reset Packet
This design ensures that a stateless reset packet is - to the extent This design ensures that a stateless reset packet is - to the extent
possible - indistinguishable from a regular packet with a short possible - indistinguishable from a regular packet with a short
header. header.
The message consists of a header octet, followed by an arbitrary The message consists of a header octet, followed by an arbitrary
number of random octets, followed by a Stateless Reset Token. number of random octets, followed by a Stateless Reset Token.
A stateless reset will be interpreted by a recipient as a packet with A stateless reset will be interpreted by a recipient as a packet with
a short header. For the packet to appear as valid, the Random Octets a short header. For the packet to appear as valid, the Random Octets
skipping to change at page 65, line 11 skipping to change at page 55, line 25
Because the stateless reset token is not available until connection Because the stateless reset token is not available until connection
establishment is complete or near completion, ignoring an unknown establishment is complete or near completion, ignoring an unknown
packet with a long header might be more effective. packet with a long header might be more effective.
An endpoint cannot determine the Source Connection ID from a packet An endpoint cannot determine the Source Connection ID from a packet
with a short header, therefore it cannot set the Destination with a short header, therefore it cannot set the Destination
Connection ID in the stateless reset packet. The Destination Connection ID in the stateless reset packet. The Destination
Connection ID will therefore differ from the value used in previous Connection ID will therefore differ from the value used in previous
packets. A random Destination Connection ID makes the connection ID packets. A random Destination Connection ID makes the connection ID
appear to be the result of moving to a new connection ID that was appear to be the result of moving to a new connection ID that was
provided using a NEW_CONNECTION_ID frame (Section 7.13). provided using a NEW_CONNECTION_ID frame (Section 19.12).
Using a randomized connection ID results in two problems: Using a randomized connection ID results in two problems:
o The packet might not reach the peer. If the Destination o The packet might not reach the peer. If the Destination
Connection ID is critical for routing toward the peer, then this Connection ID is critical for routing toward the peer, then this
packet could be incorrectly routed. This might also trigger packet could be incorrectly routed. This might also trigger
another Stateless Reset in response, see Section 6.13.4.3. A another Stateless Reset in response, see Section 10.4.3. A
Stateless Reset that is not correctly routed is ineffective in Stateless Reset that is not correctly routed is ineffective in
causing errors to be quickly detected and recovered. In this causing errors to be quickly detected and recovered. In this
case, endpoints will need to rely on other methods - such as case, endpoints will need to rely on other methods - such as
timers - to detect that the connection has failed. timers - to detect that the connection has failed.
o The randomly generated connection ID can be used by entities other o The randomly generated connection ID can be used by entities other
than the peer to identify this as a potential stateless reset. An than the peer to identify this as a potential stateless reset. An
endpoint that occasionally uses different connection IDs might endpoint that occasionally uses different connection IDs might
introduce some uncertainty about this. introduce some uncertainty about this.
skipping to change at page 65, line 45 skipping to change at page 56, line 10
sufficient state to do so. sufficient state to do so.
This stateless reset design is specific to QUIC version 1. An This stateless reset design is specific to QUIC version 1. An
endpoint that supports multiple versions of QUIC needs to generate a endpoint that supports multiple versions of QUIC needs to generate a
stateless reset that will be accepted by peers that support any stateless reset that will be accepted by peers that support any
version that the endpoint might support (or might have supported version that the endpoint might support (or might have supported
prior to losing state). Designers of new versions of QUIC need to be prior to losing state). Designers of new versions of QUIC need to be
aware of this and either reuse this design, or use a portion of the aware of this and either reuse this design, or use a portion of the
packet other than the last 16 octets for carrying data. packet other than the last 16 octets for carrying data.
6.13.4.1. Detecting a Stateless Reset 10.4.1. Detecting a Stateless Reset
An endpoint detects a potential stateless reset when a packet with a An endpoint detects a potential stateless reset when a packet with a
short header either cannot be decrypted or is marked as a duplicate short header either cannot be decrypted or is marked as a duplicate
packet. The endpoint then compares the last 16 octets of the packet packet. The endpoint then compares the last 16 octets of the packet
with the Stateless Reset Token provided by its peer, either in a with the Stateless Reset Token provided by its peer, either in a
NEW_CONNECTION_ID frame or the server's transport parameters. If NEW_CONNECTION_ID frame or the server's transport parameters. If
these values are identical, the endpoint MUST enter the draining these values are identical, the endpoint MUST enter the draining
period and not send any further packets on this connection. If the period and not send any further packets on this connection. If the
comparison fails, the packet can be discarded. comparison fails, the packet can be discarded.
6.13.4.2. Calculating a Stateless Reset Token 10.4.2. Calculating a Stateless Reset Token
The stateless reset token MUST be difficult to guess. In order to The stateless reset token MUST be difficult to guess. In order to
create a Stateless Reset Token, an endpoint could randomly generate create a Stateless Reset Token, an endpoint could randomly generate
[RFC4086] a secret for every connection that it creates. However, [RFC4086] a secret for every connection that it creates. However,
this presents a coordination problem when there are multiple this presents a coordination problem when there are multiple
instances in a cluster or a storage problem for an endpoint that instances in a cluster or a storage problem for an endpoint that
might lose state. Stateless reset specifically exists to handle the might lose state. Stateless reset specifically exists to handle the
case where state is lost, so this approach is suboptimal. case where state is lost, so this approach is suboptimal.
A single static key can be used across all connections to the same A single static key can be used across all connections to the same
endpoint by generating the proof using a second iteration of a endpoint by generating the proof using a second iteration of a
preimage-resistant function that takes a static key and the preimage-resistant function that takes a static key and the
connection ID chosen by the endpoint (see Section 6.1) as input. An connection ID chosen by the endpoint (see Section 5.1) as input. An
endpoint could use HMAC [RFC2104] (for example, HMAC(static_key, endpoint could use HMAC [RFC2104] (for example, HMAC(static_key,
connection_id)) or HKDF [RFC5869] (for example, using the static key connection_id)) or HKDF [RFC5869] (for example, using the static key
as input keying material, with the connection ID as salt). The as input keying material, with the connection ID as salt). The
output of this function is truncated to 16 octets to produce the output of this function is truncated to 16 octets to produce the
Stateless Reset Token for that connection. Stateless Reset Token for that connection.
An endpoint that loses state can use the same method to generate a An endpoint that loses state can use the same method to generate a
valid Stateless Reset Token. The connection ID comes from the packet valid Stateless Reset Token. The connection ID comes from the packet
that the endpoint receives. that the endpoint receives.
This design relies on the peer always sending a connection ID in its This design relies on the peer always sending a connection ID in its
packets so that the endpoint can use the connection ID from a packet packets so that the endpoint can use the connection ID from a packet
to reset the connection. An endpoint that uses this design MUST to reset the connection. An endpoint that uses this design MUST
either use the same connection ID length for all connections or either use the same connection ID length for all connections or
encode the length of the connection ID such that it can be recovered encode the length of the connection ID such that it can be recovered
without state. In addition, it MUST NOT provide a zero-length without state. In addition, it cannot provide a zero-length
connection ID. connection ID.
Revealing the Stateless Reset Token allows any entity to terminate Revealing the Stateless Reset Token allows any entity to terminate
the connection, so a value can only be used once. This method for the connection, so a value can only be used once. This method for
choosing the Stateless Reset Token means that the combination of choosing the Stateless Reset Token means that the combination of
connection ID and static key cannot occur for another connection. A connection ID and static key cannot occur for another connection. A
denial of service attack is possible if the same connection ID is denial of service attack is possible if the same connection ID is
used by instances that share a static key, or if an attacker can used by instances that share a static key, or if an attacker can
cause a packet to be routed to an instance that has no state but the cause a packet to be routed to an instance that has no state but the
same static key (see Section 12.8). A connection ID from a same static key (see Section 21.8). A connection ID from a
connection that is reset by revealing the Stateless Reset Token connection that is reset by revealing the Stateless Reset Token
cannot be reused for new connections at nodes that share a static cannot be reused for new connections at nodes that share a static
key. key.
Note that Stateless Reset packets do not have any cryptographic Note that Stateless Reset packets do not have any cryptographic
protection. protection.
6.13.4.3. Looping 10.4.3. Looping
The design of a Stateless Reset is such that it is indistinguishable The design of a Stateless Reset is such that it is indistinguishable
from a valid packet. This means that a Stateless Reset might trigger from a valid packet. This means that a Stateless Reset might trigger
the sending of a Stateless Reset in response, which could lead to the sending of a Stateless Reset in response, which could lead to
infinite exchanges. infinite exchanges.
An endpoint MUST ensure that every Stateless Reset that it sends is An endpoint MUST ensure that every Stateless Reset that it sends is
smaller than the packet which triggered it, unless it maintains state smaller than the packet which triggered it, unless it maintains state
sufficient to prevent looping. In the event of a loop, this results sufficient to prevent looping. In the event of a loop, this results
in packets eventually being too small to trigger a response. in packets eventually being too small to trigger a response.
skipping to change at page 67, line 39 skipping to change at page 58, line 5
observer that it is a Stateless Reset. Conversely, refusing to send observer that it is a Stateless Reset. Conversely, refusing to send
a Stateless Reset in response to a small packet might result in a Stateless Reset in response to a small packet might result in
Stateless Reset not being useful in detecting cases of broken Stateless Reset not being useful in detecting cases of broken
connections where only very small packets are sent; such failures connections where only very small packets are sent; such failures
might only be detected by other means, such as timers. might only be detected by other means, such as timers.
An endpoint can increase the odds that a packet will trigger a An endpoint can increase the odds that a packet will trigger a
Stateless Reset if it cannot be processed by padding it to at least Stateless Reset if it cannot be processed by padding it to at least
38 octets. 38 octets.
7. Frame Types and Formats 11. Error Handling
As described in Section 5, packets contain one or more frames. This An endpoint that detects an error SHOULD signal the existence of that
section describes the format and semantics of the core QUIC frame error to its peer. Both transport-level and application-level errors
can affect an entire connection (see Section 11.1), while only
application-level errors can be isolated to a single stream (see
Section 11.2).
The most appropriate error code (Section 20) SHOULD be included in
the frame that signals the error. Where this specification
identifies error conditions, it also identifies the error code that
is used.
A stateless reset (Section 10.4) is not suitable for any error that
can be signaled with a CONNECTION_CLOSE, APPLICATION_CLOSE, or
RST_STREAM frame. A stateless reset MUST NOT be used by an endpoint
that has the state necessary to send a frame on the connection.
11.1. Connection Errors
Errors that result in the connection being unusable, such as an
obvious violation of protocol semantics or corruption of state that
affects an entire connection, MUST be signaled using a
CONNECTION_CLOSE or APPLICATION_CLOSE frame (Section 19.3,
Section 19.4). An endpoint MAY close the connection in this manner
even if the error only affects a single stream.
Application protocols can signal application-specific protocol errors
using the APPLICATION_CLOSE frame. Errors that are specific to the
transport, including all those described in this document, are
carried in a CONNECTION_CLOSE frame. Other than the type of error
code they carry, these frames are identical in format and semantics.
A CONNECTION_CLOSE or APPLICATION_CLOSE frame could be sent in a
packet that is lost. An endpoint SHOULD be prepared to retransmit a
packet containing either frame type if it receives more packets on a
terminated connection. Limiting the number of retransmissions and
the time over which this final packet is sent limits the effort
expended on terminated connections.
An endpoint that chooses not to retransmit packets containing
CONNECTION_CLOSE or APPLICATION_CLOSE risks a peer missing the first
such packet. The only mechanism available to an endpoint that
continues to receive data for a terminated connection is to use the
stateless reset process (Section 10.4).
An endpoint that receives an invalid CONNECTION_CLOSE or
APPLICATION_CLOSE frame MUST NOT signal the existence of the error to
its peer.
11.2. Stream Errors
If an application-level error affects a single stream, but otherwise
leaves the connection in a recoverable state, the endpoint can send a
RST_STREAM frame (Section 19.2) with an appropriate error code to
terminate just the affected stream.
Other than STOPPING (Section 3.5), RST_STREAM MUST be instigated by
the application and MUST carry an application error code. Resetting
a stream without knowledge of the application protocol could cause
the protocol to enter an unrecoverable state. Application protocols
might require certain streams to be reliably delivered in order to
guarantee consistent state between endpoints.
12. Packets and Frames
QUIC endpoints communicate by exchanging packets. Packets are
carried in UDP datagrams (see Section 12.2) and have confidentiality
and integrity protection (see Section 12.1).
This version of QUIC uses the long packet header (see Section 17.2)
during connection establishment and the short header (see
Section 17.3) once 1-RTT keys have been established.
Packets that carry the long header are Initial Section 17.5, Retry
Section 17.7, Handshake Section 17.6, and 0-RTT Protected packets
Section 12.1.
Packets with the short header are designed for minimal overhead and
are used after a connection is established.
Version negotiation uses a packet with a special format (see
Section 17.4).
12.1. Protected Packets
All QUIC packets except Version Negotiation and Retry packets use
authenticated encryption with additional data (AEAD) [RFC5119] to
provide confidentiality and integrity protection. Details of packet
protection are found in [QUIC-TLS]; this section includes an overview
of the process.
Initial packets are protected using keys that are statically derived.
This packet protection is not effective confidentiality protection,
it only exists to ensure that the sender of the packet is on the
network path. Any entity that receives the Initial packet from a
client can recover the keys necessary to remove packet protection or
to generate packets that will be successfully authenticated.
All other packets are protected with keys derived from the
cryptographic handshake. The type of the packet from the long header
or key phase from the short header are used to identify which
encryption level - and therefore the keys - that are used. Packets
protected with 0-RTT and 1-RTT keys are expected to have
confidentiality and data origin authentication; the cryptographic
handshake ensures that only the communicating endpoints receive the
corresponding keys.
The packet number field contains a packet number, which has
additional confidentiality protection that is applied after packet
protection is applied (see [QUIC-TLS] for details). The underlying
packet number increases with each packet sent, see Section 12.3 for
details.
12.2. Coalescing Packets
A sender can coalesce multiple QUIC packets into one UDP datagram.
This can reduce the number of UDP datagrams needed to complete the
cryptographic handshake and starting sending data. Receivers MUST be
able to process coalesced packets.
Coalescing packets in order of increasing encryption levels (Initial,
0-RTT, Handshake, 1-RTT) makes it more likely the receiver will be
able to process all the packets in a single pass. A packet with a
short header does not include a length, so it will always be the last
packet included in a UDP datagram.
Senders MUST NOT coalesce QUIC packets for different connections into
a single UDP datagram. Receivers SHOULD ignore any subsequent
packets with a different Destination Connection ID than the first
packet in the datagram.
Every QUIC packet that is coalesced into a single UDP datagram is
separate and complete. Though the values of some fields in the
packet header might be redundant, no fields are omitted. The
receiver of coalesced QUIC packets MUST individually process each
QUIC packet and separately acknowledge them, as if they were received
as the payload of different UDP datagrams. For example, if
decryption fails (because the keys are not available or any other
reason) or the packet is of an unknown type, the receiver MAY either
discard or buffer the packet for later processing and MUST attempt to
process the remaining packets.
Retry packets (Section 17.7), Version Negotiation packets
(Section 17.4), and packets with a short header cannot be followed by
other packets in the same UDP datagram.
12.3. Packet Numbers
The packet number is an integer in the range 0 to 2^62-1. Where
present, packet numbers are encoded as a variable-length integer (see
Section 16). This number is used in determining the cryptographic
nonce for packet protection. Each endpoint maintains a separate
packet number for sending and receiving.
Version Negotiation (Section 17.4) and Retry Section 17.7 packets do
not include a packet number.
Packet numbers are divided into 3 spaces in QUIC:
o Initial space: All Initial packets Section 17.5 are in this space.
o Handshake space: All Handshake packets Section 17.6 are in this
space.
o Application data space: All 0-RTT and 1-RTT encrypted packets
Section 12.1 are in this space.
As described in [QUIC-TLS], each packet type uses different
protection keys.
Conceptually, a packet number space is the context in which a packet
can be processed and acknowledged. Initial packets can only be sent
with Initial packet protection keys and acknowledged in packets which
are also Initial packets. Similarly, Handshake packets are sent at
the Handshake encryption level and can only be acknowledged in
Handshake packets.
This enforces cryptographic separation between the data sent in the
different packet sequence number spaces. Each packet number space
starts at packet number 0. Subsequent packets sent in the same
packet number space MUST increase the packet number by at least one.
0-RTT and 1-RTT data exist in the same packet number space to make
loss recovery algorithms easier to implement between the two packet
types. types.
7.1. Variable-Length Integer Encoding A QUIC endpoint MUST NOT reuse a packet number within the same packet
number space in one connection (that is, under the same cryptographic
keys). If the packet number for sending reaches 2^62 - 1, the sender
MUST close the connection without sending a CONNECTION_CLOSE frame or
any further packets; an endpoint MAY send a Stateless Reset
(Section 10.4) in response to further packets that it receives.
QUIC frames commonly use a variable-length encoding for non-negative A receiver MUST discard a newly unprotected packet unless it is
integer values. This encoding ensures that smaller integer values certain that it has not processed another packet with the same packet
need fewer octets to encode. number from the same packet number space. Duplicate suppression MUST
happen after removing packet protection for the reasons described in
Section 9.3 of [QUIC-TLS]. An efficient algorithm for duplicate
suppression can be found in Section 3.4.3 of [RFC2406].
Packet number encoding at a sender and decoding at a receiver are
described in Section 17.1.
12.4. Frames and Frame Types
The payload of QUIC packets, after removing packet protection,
commonly consists of a sequence of frames, as shown in Figure 8.
Version Negotiation, Stateless Reset, and Retry packets do not
contain frames.
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Frame 1 (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Frame 2 (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Frame N (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 8: QUIC Payload
QUIC payloads MUST contain at least one frame, and MAY contain
multiple frames and multiple frame types.
Frames MUST fit within a single QUIC packet and MUST NOT span a QUIC
packet boundary. Each frame begins with a Frame Type, indicating its
type, followed by additional type-dependent fields:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Frame Type (i) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type-Dependent Fields (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 9: Generic Frame Layout
The frame types defined in this specification are listed in Table 3.
The Frame Type in STREAM frames is used to carry other frame-specific
flags. For all other frames, the Frame Type field simply identifies
the frame. These frames are explained in more detail in Section 19.
+-------------+----------------------+---------------+
| Type Value | Frame Type Name | Definition |
+-------------+----------------------+---------------+
| 0x00 | PADDING | Section 19.1 |
| | | |
| 0x01 | RST_STREAM | Section 19.2 |
| | | |
| 0x02 | CONNECTION_CLOSE | Section 19.3 |
| | | |
| 0x03 | APPLICATION_CLOSE | Section 19.4 |
| | | |
| 0x04 | MAX_DATA | Section 19.5 |
| | | |
| 0x05 | MAX_STREAM_DATA | Section 19.6 |
| | | |
| 0x06 | MAX_STREAM_ID | Section 19.7 |
| | | |
| 0x07 | PING | Section 19.8 |
| | | |
| 0x08 | BLOCKED | Section 19.9 |
| | | |
| 0x09 | STREAM_BLOCKED | Section 19.10 |
| | | |
| 0x0a | STREAM_ID_BLOCKED | Section 19.11 |
| | | |
| 0x0b | NEW_CONNECTION_ID | Section 19.12 |
| | | |
| 0x0c | STOP_SENDING | Section 19.14 |
| | | |
| 0x0d | RETIRE_CONNECTION_ID | Section 19.13 |
| | | |
| 0x0e | PATH_CHALLENGE | Section 19.16 |
| | | |
| 0x0f | PATH_RESPONSE | Section 19.17 |
| | | |
| 0x10 - 0x17 | STREAM | Section 19.19 |
| | | |
| 0x18 | CRYPTO | Section 19.20 |
| | | |
| 0x19 | NEW_TOKEN | Section 19.18 |
| | | |
| 0x1a - 0x1b | ACK | Section 19.15 |
+-------------+----------------------+---------------+
Table 3: Frame Types
All QUIC frames are idempotent. That is, a valid frame does not
cause undesirable side effects or errors when received more than
once.
The Frame Type field uses a variable length integer encoding (see
Section 16) with one exception. To ensure simple and efficient
implementations of frame parsing, a frame type MUST use the shortest
possible encoding. Though a two-, four- or eight-octet encoding of
the frame types defined in this document is possible, the Frame Type
field for these frames is encoded on a single octet. For instance,
though 0x4007 is a legitimate two-octet encoding for a variable-
length integer with a value of 7, PING frames are always encoded as a
single octet with the value 0x07. An endpoint MUST treat the receipt
of a frame type that uses a longer encoding than necessary as a
connection error of type PROTOCOL_VIOLATION.
13. Packetization and Reliability
A sender bundles one or more frames in a QUIC packet (see
Section 12.4).
A sender can minimize per-packet bandwidth and computational costs by
bundling as many frames as possible within a QUIC packet. A sender
MAY wait for a short period of time to bundle multiple frames before
sending a packet that is not maximally packed, to avoid sending out
large numbers of small packets. An implementation may use knowledge
about application sending behavior or heuristics to determine whether
and for how long to wait. This waiting period is an implementation
decision, and an implementation should be careful to delay
conservatively, since any delay is likely to increase application-
visible latency.
Stream multiplexing is achieved by interleaving STREAM frames from
multiple streams into one or more QUIC packets. A single QUIC packet
can include multiple STREAM frames from one or more streams.
One of the benefits of QUIC is avoidance of head-of-line blocking
across multiple streams. When a packet loss occurs, only streams
with data in that packet are blocked waiting for a retransmission to
be received, while other streams can continue making progress. Note
that when data from multiple streams is bundled into a single QUIC
packet, loss of that packet blocks all those streams from making
progress. Implementations are advised to bundle as few streams as
necessary in outgoing packets without losing transmission efficiency
to underfilled packets.
13.1. Packet Processing and Acknowledgment
A packet MUST NOT be acknowledged until packet protection has been
successfully removed and all frames contained in the packet have been
processed. For STREAM frames, this means the data has been enqueued
in preparation to be received by the application protocol, but it
does not require that data is delivered and consumed.
Once the packet has been fully processed, a receiver acknowledges
receipt by sending one or more ACK frames containing the packet
number of the received packet.
13.1.1. Sending ACK Frames
To avoid creating an indefinite feedback loop, an endpoint MUST NOT
send an ACK frame in response to a packet containing only ACK or
PADDING frames, even if there are packet gaps which precede the
received packet. The endpoint MUST however acknowledge packets
containing only ACK or PADDING frames when sending ACK frames in
response to other packets.
While PADDING frames do not elicit an ACK frame from a receiver, they
are considered to be in flight for congestion control purposes
[QUIC-RECOVERY]. Sending only PADDING frames might cause the sender
to become limited by the congestion controller (as described in
[QUIC-RECOVERY]) with no acknowledgments forthcoming from the
receiver. Therefore, a sender should ensure that other frames are
sent in addition to PADDING frames to elicit acknowledgments from the
receiver.
An endpoint MUST NOT send more than one packet containing only an ACK
frame per received packet that contains frames other than ACK and
PADDING frames.
The receiver's delayed acknowledgment timer SHOULD NOT exceed the
current RTT estimate or the value it indicates in the "max_ack_delay"
transport parameter. This ensures an acknowledgment is sent at least
once per RTT when packets needing acknowledgement are received. The
sender can use the receiver's "max_ack_delay" value in determining
timeouts for timer-based retransmission.
Strategies and implications of the frequency of generating
acknowledgments are discussed in more detail in [QUIC-RECOVERY].
To limit ACK Blocks to those that have not yet been received by the
sender, the receiver SHOULD track which ACK frames have been
acknowledged by its peer. Once an ACK frame has been acknowledged,
the packets it acknowledges SHOULD NOT be acknowledged again.
Because ACK frames are not sent in response to ACK-only packets, a
receiver that is only sending ACK frames will only receive
acknowledgements for its packets if the sender includes them in
packets with non-ACK frames. A sender SHOULD bundle ACK frames with
other frames when possible.
To limit receiver state or the size of ACK frames, a receiver MAY
limit the number of ACK Blocks it sends. A receiver can do this even
without receiving acknowledgment of its ACK frames, with the
knowledge this could cause the sender to unnecessarily retransmit
some data. Standard QUIC [QUIC-RECOVERY] algorithms declare packets
lost after sufficiently newer packets are acknowledged. Therefore,
the receiver SHOULD repeatedly acknowledge newly received packets in
preference to packets received in the past.
13.1.2. ACK Frames and Packet Protection
ACK frames MUST only be carried in a packet that has the same packet
number space as the packet being ACKed (see Section 12.1). For
instance, packets that are protected with 1-RTT keys MUST be
acknowledged in packets that are also protected with 1-RTT keys.
Packets that a client sends with 0-RTT packet protection MUST be
acknowledged by the server in packets protected by 1-RTT keys. This
can mean that the client is unable to use these acknowledgments if
the server cryptographic handshake messages are delayed or lost.
Note that the same limitation applies to other data sent by the
server protected by the 1-RTT keys.
Endpoints SHOULD send acknowledgments for packets containing CRYPTO
frames with a reduced delay; see Section 4.3.1 of [QUIC-RECOVERY].
13.2. Retransmission of Information
QUIC packets that are determined to be lost are not retransmitted
whole. The same applies to the frames that are contained within lost
packets. Instead, the information that might be carried in frames is
sent again in new frames as needed.
New frames and packets are used to carry information that is
determined to have been lost. In general, information is sent again
when a packet containing that information is determined to be lost
and sending ceases when a packet containing that information is
acknowledged.
o Data sent in CRYPTO frames is retransmitted according to the rules
in [QUIC-RECOVERY], until all data has been acknowledged.
o Application data sent in STREAM frames is retransmitted in new
STREAM frames unless the endpoint has sent a RST_STREAM for that
stream. Once an endpoint sends a RST_STREAM frame, no further
STREAM frames are needed.
o The most recent set of acknowledgments are sent in ACK frames. An
ACK frame SHOULD contain all unacknowledged acknowledgments, as
described in Section 13.1.1.
o Cancellation of stream transmission, as carried in a RST_STREAM
frame, is sent until acknowledged or until all stream data is
acknowledged by the peer (that is, either the "Reset Recvd" or
"Data Recvd" state is reached on the send stream). The content of
a RST_STREAM frame MUST NOT change when it is sent again.
o Similarly, a request to cancel stream transmission, as encoded in
a STOP_SENDING frame, is sent until the receive stream enters
either a "Data Recvd" or "Reset Recvd" state, see Section 3.5.
o Connection close signals, including those that use
CONNECTION_CLOSE and APPLICATION_CLOSE frames, are not sent again
when packet loss is detected, but as described in Section 10.
o The current connection maximum data is sent in MAX_DATA frames.
An updated value is sent in a MAX_DATA frame if the packet
containing the most recently sent MAX_DATA frame is declared lost,
or when the endpoint decides to update the limit. Care is
necessary to avoid sending this frame too often as the limit can
increase frequently and cause an unnecessarily large number of
MAX_DATA frames to be sent.
o The current maximum stream data offset is sent in MAX_STREAM_DATA
frames. Like MAX_DATA, an updated value is sent when the packet
containing the most recent MAX_STREAM_DATA frame for a stream is
lost or when the limit is updated, with care taken to prevent the
frame from being sent too often. An endpoint SHOULD stop sending
MAX_STREAM_DATA frames when the receive stream enters a "Size
Known" state.
o The maximum stream ID for a stream of a given type is sent in
MAX_STREAM_ID frames. Like MAX_DATA, an updated value is sent
when a packet containing the most recent MAX_STREAM_ID for a
stream type frame is declared lost or when the limit is updated,
with care taken to prevent the frame from being sent too often.
o Blocked signals are carried in BLOCKED, STREAM_BLOCKED, and
STREAM_ID_BLOCKED frames. BLOCKED streams have connection scope,
STREAM_BLOCKED frames have stream scope, and STREAM_ID_BLOCKED
frames are scoped to a specific stream type. New frames are sent
if packets containing the most recent frame for a scope is lost,
but only while the endpoint is blocked on the corresponding limit.
These frames always include the limit that is causing blocking at
the time that they are transmitted.
o A liveness or path validation check using PATH_CHALLENGE frames is
sent periodically until a matching PATH_RESPONSE frame is received
or until there is no remaining need for liveness or path
validation checking. PATH_CHALLENGE frames include a different
payload each time they are sent.
o Responses to path validation using PATH_RESPONSE frames are sent
just once. A new PATH_CHALLENGE frame will be sent if another
PATH_RESPONSE frame is needed.
o New connection IDs are sent in NEW_CONNECTION_ID frames and
retransmitted if the packet containing them is lost.
Retransmissions of this frame carry the same sequence number
value. Likewise, retired connection IDs are sent in
RETIRE_CONNECTION_ID frames and retransmitted if the packet
containing them is lost.
o PADDING frames contain no information, so lost PADDING frames do
not require repair.
Upon detecting losses, a sender MUST take appropriate congestion
control action. The details of loss detection and congestion control
are described in [QUIC-RECOVERY].
13.3. Explicit Congestion Notification
QUIC endpoints use Explicit Congestion Notification (ECN) [RFC3168]
to detect and respond to network congestion. ECN allows a network
node to indicate congestion in the network by setting a codepoint in
the IP header of a packet instead of dropping it. Endpoints react to
congestion by reducing their sending rate in response, as described
in [QUIC-RECOVERY].
To use ECN, QUIC endpoints first determine whether a path supports
ECN marking and the peer is able to access the ECN codepoint in the
IP header. A network path does not support ECN if ECN marked packets
get dropped or ECN markings are rewritten on the path. An endpoint
verifies the path, both during connection establishment and when
migrating to a new path (see Section 9).
13.3.1. ECN Counters
On receiving a packet with an ECT or CE codepoint, an endpoint that
can access the IP ECN codepoints increases the corresponding ECT(0),
ECT(1), or CE count, and includes these counters in subsequent (see
Section 13.1) ACK frames (see Section 19.15).
A packet detected by a receiver as a duplicate does not affect the
receiver's local ECN codepoint counts; see (Section 21.7) for
relevant security concerns.
If an endpoint receives a packet without an ECT or CE codepoint, it
responds per Section 13.1 with an ACK frame. If an endpoint does not
have access to received ECN codepoints, it acknowledges received
packets per Section 13.1 with an ACK frame.
13.3.2. ECN Verification
Each endpoint independently verifies and enables use of ECN by
setting the IP header ECN codepoint to ECN Capable Transport (ECT)
for the path from it to the other peer. Even if ECN is not used on
the path to the peer, the endpoint MUST provide feedback about ECN
markings received (if accessible).
To verify both that a path supports ECN and the peer can provide ECN
feedback, an endpoint MUST set the ECT(0) codepoint in the IP header
of all outgoing packets [RFC8311].
If an ECT codepoint set in the IP header is not corrupted by a
network device, then a received packet contains either the codepoint
sent by the peer or the Congestion Experienced (CE) codepoint set by
a network device that is experiencing congestion.
If a packet sent with an ECT codepoint is newly acknowledged by the
peer in an ACK frame without ECN feedback, the endpoint stops setting
ECT codepoints in subsequent packets, with the expectation that
either the network or the peer no longer supports ECN.
To protect the connection from arbitrary corruption of ECN codepoints
by the network, an endpoint verifies the following when an ACK frame
is received:
o The increase in ECT(0) and ECT(1) counters MUST be at least the
number of packets newly acknowledged that were sent with the
corresponding codepoint.
o The total increase in ECT(0), ECT(1), and CE counters reported in
the ACK frame MUST be at least the total number of packets newly
acknowledged in this ACK frame.
An endpoint could miss acknowledgements for a packet when ACK frames
are lost. It is therefore possible for the total increase in ECT(0),
ECT(1), and CE counters to be greater than the number of packets
acknowledged in an ACK frame. When this happens, the local reference
counts MUST be increased to match the counters in the ACK frame.
Upon successful verification, an endpoint continues to set ECT
codepoints in subsequent packets with the expectation that the path
is ECN-capable.
If verification fails, then the endpoint ceases setting ECT
codepoints in subsequent packets with the expectation that either the
network or the peer does not support ECN.
If an endpoint sets ECT codepoints on outgoing packets and encounters
a retransmission timeout due to the absence of acknowledgments from
the peer (see [QUIC-RECOVERY]), or if an endpoint has reason to
believe that a network element might be corrupting ECN codepoints,
the endpoint MAY cease setting ECT codepoints in subsequent packets.
Doing so allows the connection to traverse network elements that drop
or corrupt ECN codepoints in the IP header.
14. Packet Size
The QUIC packet size includes the QUIC header and integrity check,
but not the UDP or IP header.
Clients MUST ensure that the first Initial packet they send is sent
in a UDP datagram that is at least 1200 octets. Padding the Initial
packet or including a 0-RTT packet in the same datagram are ways to
meet this requirement. Sending a UDP datagram of this size ensures
that the network path supports a reasonable Maximum Transmission Unit
(MTU), and helps reduce the amplitude of amplification attacks caused
by server responses toward an unverified client address, see
Section 8.
The payload of a UDP datagram carrying the Initial packet MUST be
expanded to at least 1200 octets, by adding PADDING frames to the
Initial packet and/or by combining the Initial packet with a 0-RTT
packet (see Section 12.2).
The datagram containing the first Initial packet from a client MAY
exceed 1200 octets if the client believes that the Path Maximum
Transmission Unit (PMTU) supports the size that it chooses.
A server MAY send a CONNECTION_CLOSE frame with error code
PROTOCOL_VIOLATION in response to the first Initial packet it
receives from a client if the UDP datagram is smaller than 1200
octets. It MUST NOT send any other frame type in response, or
otherwise behave as if any part of the offending packet was processed
as valid.
The server MUST also limit the number of bytes it sends before
validating the address of the client, see Section 8.
14.1. Path Maximum Transmission Unit
The Path Maximum Transmission Unit (PMTU) is the maximum size of the
entire IP header, UDP header, and UDP payload. The UDP payload
includes the QUIC packet header, protected payload, and any
authentication fields.
All QUIC packets SHOULD be sized to fit within the estimated PMTU to
avoid IP fragmentation or packet drops. To optimize bandwidth
efficiency, endpoints SHOULD use Packetization Layer PMTU Discovery
([PLPMTUD]). Endpoints MAY use PMTU Discovery ([PMTUDv4], [PMTUDv6])
for detecting the PMTU, setting the PMTU appropriately, and storing
the result of previous PMTU determinations.
In the absence of these mechanisms, QUIC endpoints SHOULD NOT send IP
packets larger than 1280 octets. Assuming the minimum IP header
size, this results in a QUIC packet size of 1232 octets for IPv6 and
1252 octets for IPv4. Some QUIC implementations MAY be more
conservative in computing allowed QUIC packet size given unknown
tunneling overheads or IP header options.
QUIC endpoints that implement any kind of PMTU discovery SHOULD
maintain an estimate for each combination of local and remote IP
addresses. Each pairing of local and remote addresses could have a
different maximum MTU in the path.
QUIC depends on the network path supporting an MTU of at least 1280
octets. This is the IPv6 minimum MTU and therefore also supported by
most modern IPv4 networks. An endpoint MUST NOT reduce its MTU below
this number, even if it receives signals that indicate a smaller
limit might exist.
If a QUIC endpoint determines that the PMTU between any pair of local
and remote IP addresses has fallen below 1280 octets, it MUST
immediately cease sending QUIC packets on the affected path. This
could result in termination of the connection if an alternative path
cannot be found.
14.1.1. IPv4 PMTU Discovery
Traditional ICMP-based path MTU discovery in IPv4 [PMTUDv4] is
potentially vulnerable to off-path attacks that successfully guess
the IP/port 4-tuple and reduce the MTU to a bandwidth-inefficient
value. TCP connections mitigate this risk by using the (at minimum)
8 bytes of transport header echoed in the ICMP message to validate
the TCP sequence number as valid for the current connection.
However, as QUIC operates over UDP, in IPv4 the echoed information
could consist only of the IP and UDP headers, which usually has
insufficient entropy to mitigate off-path attacks.
As a result, endpoints that implement PMTUD in IPv4 SHOULD take steps
to mitigate this risk. For instance, an application could:
o Set the IPv4 Don't Fragment (DF) bit on a small proportion of
packets, so that most invalid ICMP messages arrive when there are
no DF packets outstanding, and can therefore be identified as
spurious.
o Store additional information from the IP or UDP headers from DF
packets (for example, the IP ID or UDP checksum) to further
authenticate incoming Datagram Too Big messages.
o Any reduction in PMTU due to a report contained in an ICMP packet
is provisional until QUIC's loss detection algorithm determines
that the packet is actually lost.
14.2. Special Considerations for Packetization Layer PMTU Discovery
The PADDING frame provides a useful option for PMTU probe packets.
PADDING frames generate acknowledgements, but they need not be
delivered reliably. As a result, the loss of PADDING frames in probe
packets does not require delay-inducing retransmission. However,
PADDING frames do consume congestion window, which may delay the
transmission of subsequent application data.
When implementing the algorithm in Section 7.2 of [PLPMTUD], the
initial value of search_low SHOULD be consistent with the IPv6
minimum packet size. Paths that do not support this size cannot
deliver Initial packets, and therefore are not QUIC-compliant.
Section 7.3 of [PLPMTUD] discusses trade-offs between small and large
increases in the size of probe packets. As QUIC probe packets need
not contain application data, aggressive increases in probe size
carry fewer consequences.
15. Versions
QUIC versions are identified using a 32-bit unsigned number.
The version 0x00000000 is reserved to represent version negotiation.
This version of the specification is identified by the number
0x00000001.
Other versions of QUIC might have different properties to this
version. The properties of QUIC that are guaranteed to be consistent
across all versions of the protocol are described in
[QUIC-INVARIANTS].
Version 0x00000001 of QUIC uses TLS as a cryptographic handshake
protocol, as described in [QUIC-TLS].
Versions with the most significant 16 bits of the version number
cleared are reserved for use in future IETF consensus documents.
Versions that follow the pattern 0x?a?a?a?a are reserved for use in
forcing version negotiation to be exercised. That is, any version
number where the low four bits of all octets is 1010 (in binary). A
client or server MAY advertise support for any of these reserved
versions.
Reserved version numbers will probably never represent a real
protocol; a client MAY use one of these version numbers with the
expectation that the server will initiate version negotiation; a
server MAY advertise support for one of these versions and can expect
that clients ignore the value.
[[RFC editor: please remove the remainder of this section before
publication.]]
The version number for the final version of this specification
(0x00000001), is reserved for the version of the protocol that is
published as an RFC.
Version numbers used to identify IETF drafts are created by adding
the draft number to 0xff000000. For example, draft-ietf-quic-
transport-13 would be identified as 0xff00000D.
Implementors are encouraged to register version numbers of QUIC that
they are using for private experimentation on the GitHub wiki at
<https://github.com/quicwg/base-drafts/wiki/QUIC-Versions>.
16. Variable-Length Integer Encoding
QUIC packets and frames commonly use a variable-length encoding for
non-negative integer values. This encoding ensures that smaller
integer values need fewer octets to encode.
The QUIC variable-length integer encoding reserves the two most The QUIC variable-length integer encoding reserves the two most
significant bits of the first octet to encode the base 2 logarithm of significant bits of the first octet to encode the base 2 logarithm of
the integer encoding length in octets. The integer value is encoded the integer encoding length in octets. The integer value is encoded
on the remaining bits, in network byte order. on the remaining bits, in network byte order.
This means that integers are encoded on 1, 2, 4, or 8 octets and can This means that integers are encoded on 1, 2, 4, or 8 octets and can
encode 6, 14, 30, or 62 bit values respectively. Table 4 summarizes encode 6, 14, 30, or 62 bit values respectively. Table 4 summarizes
the encoding properties. the encoding properties.
skipping to change at page 68, line 31 skipping to change at page 75, line 40
+------+--------+-------------+-----------------------+ +------+--------+-------------+-----------------------+
Table 4: Summary of Integer Encodings Table 4: Summary of Integer Encodings
For example, the eight octet sequence c2 19 7c 5e ff 14 e8 8c (in For example, the eight octet sequence c2 19 7c 5e ff 14 e8 8c (in
hexadecimal) decodes to the decimal value 151288809941952652; the hexadecimal) decodes to the decimal value 151288809941952652; the
four octet sequence 9d 7f 3e 7d decodes to 494878333; the two octet four octet sequence 9d 7f 3e 7d decodes to 494878333; the two octet
sequence 7b bd decodes to 15293; and the single octet 25 decodes to sequence 7b bd decodes to 15293; and the single octet 25 decodes to
37 (as does the two octet sequence 40 25). 37 (as does the two octet sequence 40 25).
Error codes (Section 11.3) are described using integers, but do not Error codes (Section 20) and versions Section 15 are described using
use this encoding. integers, but do not use this encoding.
7.2. PADDING Frame 17. Packet Formats
All numeric values are encoded in network byte order (that is, big-
endian) and all field sizes are in bits. Hexadecimal notation is
used for describing the value of fields.
17.1. Packet Number Encoding and Decoding
Packet numbers in long and short packet headers are encoded as
follows. The number of bits required to represent the packet number
is first reduced by including only a variable number of the least
significant bits of the packet number. One or two of the most
significant bits of the first octet are then used to represent how
many bits of the packet number are provided, as shown in Table 5.
+---------------------+----------------+--------------+
| First octet pattern | Encoded Length | Bits Present |
+---------------------+----------------+--------------+
| 0b0xxxxxxx | 1 octet | 7 |
| | | |
| 0b10xxxxxx | 2 | 14 |
| | | |
| 0b11xxxxxx | 4 | 30 |
+---------------------+----------------+--------------+
Table 5: Packet Number Encodings for Packet Headers
Note that these encodings are similar to those in Section 16, but use
different values.
Finally, the encoded packet number is protected as described in
Section 5.3 of [QUIC-TLS].
The sender MUST use a packet number size able to represent more than
twice as large a range than the difference between the largest
acknowledged packet and packet number being sent. A peer receiving
the packet will then correctly decode the packet number, unless the
packet is delayed in transit such that it arrives after many higher-
numbered packets have been received. An endpoint SHOULD use a large
enough packet number encoding to allow the packet number to be
recovered even if the packet arrives after packets that are sent
afterwards.
As a result, the size of the packet number encoding is at least one
more than the base 2 logarithm of the number of contiguous
unacknowledged packet numbers, including the new packet.
For example, if an endpoint has received an acknowledgment for packet
0x6afa2f, sending a packet with a number of 0x6b2d79 requires a
packet number encoding with 14 bits or more; whereas the 30-bit
packet number encoding is needed to send a packet with a number of
0x6bc107.
At a receiver, protection of the packet number is removed prior to
recovering the full packet number. The full packet number is then
reconstructed based on the number of significant bits present, the
value of those bits, and the largest packet number received on a
successfully authenticated packet. Recovering the full packet number
is necessary to successfully remove packet protection.
Once packet number protection is removed, the packet number is
decoded by finding the packet number value that is closest to the
next expected packet. The next expected packet is the highest
received packet number plus one. For example, if the highest
successfully authenticated packet had a packet number of 0xaa82f30e,
then a packet containing a 14-bit value of 0x9b3 will be decoded as
0xaa8309b3. Example pseudo-code for packet number decoding can be
found in Appendix A.
17.2. Long Header Packet
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+
|1| Type (7) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Version (32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|DCIL(4)|SCIL(4)|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Source Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Length (i) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Packet Number (8/16/32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Payload (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 10: Long Header Packet Format
Long headers are used for packets that are sent prior to the
completion of version negotiation and establishment of 1-RTT keys.
Once both conditions are met, a sender switches to sending packets
using the short header (Section 17.3). The long form allows for
special packets - such as the Version Negotiation packet - to be
represented in this uniform fixed-length packet format. Packets that
use the long header contain the following fields:
Header Form: The most significant bit (0x80) of octet 0 (the first
octet) is set to 1 for long headers.
Long Packet Type: The remaining seven bits of octet 0 contain the
packet type. This field can indicate one of 128 packet types.
The types specified for this version are listed in Table 6.
Version: The QUIC Version is a 32-bit field that follows the Type.
This field indicates which version of QUIC is in use and
determines how the rest of the protocol fields are interpreted.
DCIL and SCIL: The octet following the version contains the lengths
of the two connection ID fields that follow it. These lengths are
encoded as two 4-bit unsigned integers. The Destination
Connection ID Length (DCIL) field occupies the 4 high bits of the
octet and the Source Connection ID Length (SCIL) field occupies
the 4 low bits of the octet. An encoded length of 0 indicates
that the connection ID is also 0 octets in length. Non-zero
encoded lengths are increased by 3 to get the full length of the
connection ID, producing a length between 4 and 18 octets
inclusive. For example, an octet with the value 0x50 describes an
8-octet Destination Connection ID and a zero-length Source
Connection ID.
Destination Connection ID: The Destination Connection ID field
follows the connection ID lengths and is either 0 octets in length
or between 4 and 18 octets. Section 7.2 describes the use of this
field in more detail.
Source Connection ID: The Source Connection ID field follows the
Destination Connection ID and is either 0 octets in length or
between 4 and 18 octets. Section 7.2 describes the use of this
field in more detail.
Length: The length of the remainder of the packet (that is, the
Packet Number and Payload fields) in octets, encoded as a
variable-length integer (Section 16).
Packet Number: The packet number field is 1, 2, or 4 octets long.
The packet number has confidentiality protection separate from
packet protection, as described in Section 5.3 of [QUIC-TLS]. The
length of the packet number field is encoded in the plaintext
packet number. See Section 17.1 for details.
Payload: The payload of the packet.
The following packet types are defined:
+------+-----------------+--------------+
| Type | Name | Section |
+------+-----------------+--------------+
| 0x7F | Initial | Section 17.5 |
| | | |
| 0x7E | Retry | Section 17.7 |
| | | |
| 0x7D | Handshake | Section 17.6 |
| | | |
| 0x7C | 0-RTT Protected | Section 12.1 |
+------+-----------------+--------------+
Table 6: Long Header Packet Types
The header form, type, connection ID lengths octet, destination and
source connection IDs, and version fields of a long header packet are
version-independent. The packet number and values for packet types
defined in Table 6 are version-specific. See [QUIC-INVARIANTS] for
details on how packets from different versions of QUIC are
interpreted.
The interpretation of the fields and the payload are specific to a
version and packet type. Type-specific semantics for this version
are described in the following sections.
The end of the packet is determined by the Length field. The Length
field covers both the Packet Number and Payload fields, both of which
are confidentiality protected and initially of unknown length. The
size of the Payload field is learned once the packet number
protection is removed. The Length field enables packet coalescing
(Section 12.2).
17.3. Short Header Packet
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+
|0|K|1|1|0|R R R|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Packet Number (8/16/32) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Protected Payload (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 11: Short Header Packet Format
The short header can be used after the version and 1-RTT keys are
negotiated. Packets that use the short header contain the following
fields:
Header Form: The most significant bit (0x80) of octet 0 is set to 0
for the short header.
Key Phase Bit: The second bit (0x40) of octet 0 indicates the key
phase, which allows a recipient of a packet to identify the packet
protection keys that are used to protect the packet. See
[QUIC-TLS] for details.
[[Editor's Note: this section should be removed and the bit
definitions changed before this draft goes to the IESG.]]
Third Bit: The third bit (0x20) of octet 0 is set to 1.
[[Editor's Note: this section should be removed and the bit
definitions changed before this draft goes to the IESG.]]
Fourth Bit: The fourth bit (0x10) of octet 0 is set to 1.
[[Editor's Note: this section should be removed and the bit
definitions changed before this draft goes to the IESG.]]
Google QUIC Demultiplexing Bit: The fifth bit (0x8) of octet 0 is
set to 0. This allows implementations of Google QUIC to
distinguish Google QUIC packets from short header packets sent by
a client because Google QUIC servers expect the connection ID to
always be present. The special interpretation of this bit SHOULD
be removed from this specification when Google QUIC has finished
transitioning to the new header format.
Reserved: The sixth, seventh, and eighth bits (0x7) of octet 0 are
reserved for experimentation. Endpoints MUST ignore these bits on
packets they receive unless they are participating in an
experiment that uses these bits. An endpoint not actively using
these bits SHOULD set the value randomly on packets they send to
protect against unwanted inference about particular values.
Destination Connection ID: The Destination Connection ID is a
connection ID that is chosen by the intended recipient of the
packet. See Section 5.1 for more details.
Packet Number: The packet number field is 1, 2, or 4 octets long.
The packet number has confidentiality protection separate from
packet protection, as described in Section 5.3 of [QUIC-TLS]. The
length of the packet number field is encoded in the plaintext
packet number. See Section 17.1 for details.
Protected Payload: Packets with a short header always include a
1-RTT protected payload.
The header form and connection ID field of a short header packet are
version-independent. The remaining fields are specific to the
selected QUIC version. See [QUIC-INVARIANTS] for details on how
packets from different versions of QUIC are interpreted.
17.4. Version Negotiation Packet
A Version Negotiation packet is inherently not version-specific, and
does not use the long packet header (see Section 17.2). Upon receipt
by a client, it will appear to be a packet using the long header, but
will be identified as a Version Negotiation packet based on the
Version field having a value of 0.
The Version Negotiation packet is a response to a client packet that
contains a version that is not supported by the server, and is only
sent by servers.
The layout of a Version Negotiation packet is:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+
|1| Unused (7) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Version (32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|DCIL(4)|SCIL(4)|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Source Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Supported Version 1 (32) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| [Supported Version 2 (32)] ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| [Supported Version N (32)] ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 12: Version Negotiation Packet
The value in the Unused field is selected randomly by the server.
The Version field of a Version Negotiation packet MUST be set to
0x00000000.
The server MUST include the value from the Source Connection ID field
of the packet it receives in the Destination Connection ID field.
The value for Source Connection ID MUST be copied from the
Destination Connection ID of the received packet, which is initially
randomly selected by a client. Echoing both connection IDs gives
clients some assurance that the server received the packet and that
the Version Negotiation packet was not generated by an off-path
attacker.
The remainder of the Version Negotiation packet is a list of 32-bit
versions which the server supports.
A Version Negotiation packet cannot be explicitly acknowledged in an
ACK frame by a client. Receiving another Initial packet implicitly
acknowledges a Version Negotiation packet.
The Version Negotiation packet does not include the Packet Number and
Length fields present in other packets that use the long header form.
Consequently, a Version Negotiation packet consumes an entire UDP
datagram.
See Section 6 for a description of the version negotiation process.
17.5. Initial Packet
An Initial packet uses long headers with a type value of 0x7F. It
carries the first CRYPTO frames sent by the client and server to
perform key exchange, and carries ACKs in either direction.
In order to prevent tampering by version-unaware middleboxes, Initial
packets are protected with connection- and version-specific keys
(Initial keys) as described in [QUIC-TLS]. This protection does not
provide confidentiality or integrity against on-path attackers, but
provides some level of protection against off-path attackers.
An Initial packet (shown in Figure 13) has two additional header
fields that are added to the Long Header before the Length field.
+-+-+-+-+-+-+-+-+
|1| 0x7f |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Version (32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|DCIL(4)|SCIL(4)|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Destination Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Source Connection ID (0/32..144) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Token Length (i) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Token (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Length (i) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Packet Number (8/16/32) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Payload (*) ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 13: Initial Packet
These fields include the token that was previously provided in a
Retry packet or NEW_TOKEN frame:
Token Length: A variable-length integer specifying the length of the
Token field, in bytes. This value is zero if no token is present.
Initial packets sent by the server MUST set the Token Length field
to zero; clients that receive an Initial packet with a non-zero
Token Length field MUST either discard the packet or generate a
connection error of type PROTOCOL_VIOLATION.
Token: The value of the token.
The client and server use the Initial packet type for any packet that
contains an initial cryptographic handshake message. This includes
all cases where a new packet containing the initial cryptographic
message needs to be created, such as the packets sent after receiving
a Version Negotiation (Section 17.4) or Retry packet (Section 17.7).