draft-ietf-v6ops-v6onbydefault-01.txt   draft-ietf-v6ops-v6onbydefault-02.txt 
Network Working Group S. Roy Network Working Group S. Roy
Internet-Draft A. Durand Internet-Draft A. Durand
Expires: August 13, 2004 J. Paugh Expires: November 5, 2004 J. Paugh
Sun Microsystems, Inc. Sun Microsystems, Inc.
February 13, 2004 May 7, 2004
Issues with Dual Stack IPv6 on by Default Issues with Dual Stack IPv6 on by Default
draft-ietf-v6ops-v6onbydefault-01.txt draft-ietf-v6ops-v6onbydefault-02.txt
Status of this Memo Status of this Memo
This document is an Internet-Draft and is in full conformance with This document is an Internet-Draft and is in full conformance with
all provisions of Section 10 of RFC2026. all provisions of Section 10 of RFC2026.
Internet-Drafts are working documents of the Internet Engineering Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that other Task Force (IETF), its areas, and its working groups. Note that other
groups may also distribute working documents as Internet-Drafts. groups may also distribute working documents as Internet-Drafts.
skipping to change at page 1, line 32 skipping to change at page 1, line 31
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."
The list of current Internet-Drafts can be accessed at http:// The list of current Internet-Drafts can be accessed at http://
www.ietf.org/ietf/1id-abstracts.txt. www.ietf.org/ietf/1id-abstracts.txt.
The list of Internet-Draft Shadow Directories can be accessed at The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html. http://www.ietf.org/shadow.html.
This Internet-Draft will expire on August 13, 2004. This Internet-Draft will expire on November 5, 2004.
Copyright Notice Copyright Notice
Copyright (C) The Internet Society (2004). All Rights Reserved. Copyright (C) The Internet Society (2004). All Rights Reserved.
Abstract Abstract
This document discusses problems that can occur when dual stack nodes This document discusses problems that can occur when dual stack nodes
that have IPv6 enabled by default are deployed in IPv4 or mixed IPv4 that have IPv6 enabled by default are deployed in IPv4 or mixed IPv4
and IPv6 environments. The problems include application connection and IPv6 environments. The problems include application connection
delays, poor connectivity, and network security. Its purpose is to delays, poor connectivity, and network insecurity. The purpose of
raise awareness of these problems so that they can be fixed or worked this memo is to raise awareness of these problems so that they can be
around. The purpose of this document is not to try to specify whether fixed or worked around, not to try to specify whether IPv6 should be
IPv6 should be enabled by default or not, but to raise awareness of enabled by default or not.
the potential issues involved.
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . 3 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. No IPv6 Router . . . . . . . . . . . . . . . . . . . . . . 3 2. No IPv6 Router . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1 Problems with Default Address Selection for IPv6 . . . . . 3 2.1 Problems with Default Address Selection for IPv6 . . . . . 3
2.2 Neighbor Discovery's On-Link Assumption Considered 2.2 Neighbor Discovery's On-Link Assumption Considered
Harmful . . . . . . . . . . . . . . . . . . . . . . . . . 5 Harmful . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.3 Transport Protocol Robustness . . . . . . . . . . . . . . 6 2.3 Transport Protocol Robustness . . . . . . . . . . . . . . 6
2.3.1 TCP Implications . . . . . . . . . . . . . . . . . . . . . 6 2.3.1 TCP Implications . . . . . . . . . . . . . . . . . . . 6
2.3.1.1 TCP Connection Termination . . . . . . . . . . . . . . . . 6 2.3.2 UDP Implications . . . . . . . . . . . . . . . . . . . 8
2.3.1.2 Asynchronous Application Notification . . . . . . . . . . 7 2.3.3 SCTP Implications . . . . . . . . . . . . . . . . . . 8
2.3.2 UDP Implications . . . . . . . . . . . . . . . . . . . . . 7 3. Other Problematic Scenarios . . . . . . . . . . . . . . . . . 9
2.3.3 SCTP Implications . . . . . . . . . . . . . . . . . . . . 8 3.1 IPv6 Network of Smaller Scope . . . . . . . . . . . . . . 9
3. Other Problematic Scenarios . . . . . . . . . . . . . . . 8 3.1.1 Alleviating the Scope Problem . . . . . . . . . . . . 10
3.1 IPv6 Network of Smaller Scope . . . . . . . . . . . . . . 8 3.2 Poor IPv6 Network Performance . . . . . . . . . . . . . . 10
3.1.1 Alleviating the Scope Problem . . . . . . . . . . . . . . 8 3.2.1 Dealing with Poor IPv6 Network Performance . . . . . . 10
3.2 Poor IPv6 Network Performance . . . . . . . . . . . . . . 8 3.3 Security . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2.1 Dealing with Poor IPv6 Network Performance . . . . . . . . 9 3.3.1 Mitigating Security Risks . . . . . . . . . . . . . . 11
3.3 Security . . . . . . . . . . . . . . . . . . . . . . . . . 9 4. Application Robustness . . . . . . . . . . . . . . . . . . . . 12
3.3.1 Mitigating Security Risks . . . . . . . . . . . . . . . . 10 5. Security Considerations . . . . . . . . . . . . . . . . . . . 12
4. Application Robustness . . . . . . . . . . . . . . . . . . 10 6. References . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5. Security Considerations . . . . . . . . . . . . . . . . . 11 6.1 Normative References . . . . . . . . . . . . . . . . . . . . 12
Normative References . . . . . . . . . . . . . . . . . . . 11 6.2 Informative References . . . . . . . . . . . . . . . . . . . 13
Informative References . . . . . . . . . . . . . . . . . . 11 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . 13
Authors' Addresses . . . . . . . . . . . . . . . . . . . . 12 A. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 14
A. Acknowledgments . . . . . . . . . . . . . . . . . . . . . 12 B. Changes from draft-ietf-v6ops-v6onbydefault-01 . . . . . . . . 14
B. Changes from draft-ietf-v6ops-v6onbydefault-00 . . . . . . 12 C. Changes from draft-ietf-v6ops-v6onbydefault-00 . . . . . . . . 14
Intellectual Property and Copyright Statements . . . . . . 14 Intellectual Property and Copyright Statements . . . . . . . . 16
1. Introduction 1. Introduction
This document specifically addresses operating system implementations This document specifically addresses operating system implementations
that implement the dual stack IPv6 model, and would ship with IPv6 that implement the dual stack IPv6 model, and would ship with IPv6
enabled by default. It addresses the case where such a system is enabled by default. It addresses the case where such systems are
installed and placed in an IPv4 only or mixed IPv4 and IPv6 installed and placed in IPv4-only or mixed IPv4 and IPv6
environment, and documents potential problems that users on such environments, and documents potential problems that users on such
systems could experience if the IPv6 connectivity is non-existent or systems could experience if the IPv6 connectivity is non-existent or
sub-optimal. The purpose of this document is not to try to specify sub-optimal. The purpose of this document is not to try to specify
whether IPv6 should be enabled by default or not, but to raise whether IPv6 should be enabled by default or not, but to raise
awareness of the potential issues involved. awareness of the potential issues involved.
It begins in Section 2 by examining problems within IPv6 This memo begins in Section 2 by examining problems within IPv6
implementations that defeat the destination address selection implementations that defeat the destination address selection
mechanism defined in [ADDRSEL] and contribute to poor IPv6 mechanism defined in [ADDRSEL] and contribute to poor IPv6
connectivity. Starting with Section 3 it then examines other issues connectivity. Starting with Section 3 it then examines other issues
that network software engineers and network and systems that network software engineers and network and systems
administrators should be aware of when deploying dual stack systems administrators should be aware of when deploying dual stack systems
with IPv6 enabled. with IPv6 enabled.
2. No IPv6 Router 2. No IPv6 Router
Consider a scenario in which a dual stack system has IPv6 enabled and Consider a scenario in which a dual stack system has IPv6 enabled and
placed on a link with no IPv6 routers. The system is using IPv6 is placed on a link with no IPv6 routers. The system is using IPv6
Stateless Address Autoconfiguration [AUTOCONF], so it only has a Stateless Address Autoconfiguration [AUTOCONF], so it only has a
link-local IPv6 address configured. It also has a single IPv4 link-local IPv6 address configured. It also has a single IPv4
address that happens to be a private address as defined in address that happens to be a private address as defined in
[PRIVADDR]. [PRIVADDR].
An application on this system is trying to communicate with a An application on this system is trying to communicate with a
destination whose name resolves to public and global IPv4 and IPv6 destination whose name resolves to public and global IPv4 and IPv6
addresses. The application uses an address resolution API that addresses. The application uses an address resolution API that
implements the destination address selection mechanism described in implements the destination address selection mechanism described in
Default Address Selection for IPv6 [ADDRSEL]. The application will Default Address Selection for IPv6 [ADDRSEL]. The application will
attempt to connect to each address returned in order until one attempt to connect to each address, in the order they were returned,
succeeds. Since the system has no off-link IPv6 routes, the optimal until one succeeds. Since the system has no off-link IPv6 routes,
scenario would be if the IPv4 addresses returned were ordered before the optimal scenario would be if the IPv4 addresses returned were
the IPv6 addresses. The following sections describe where things can ordered before the IPv6 addresses. The following sections describe
go wrong with this scenario. what things can go wrong with this scenario.
2.1 Problems with Default Address Selection for IPv6 2.1 Problems with Default Address Selection for IPv6
The Default Address Selection for IPv6 [ADDRSEL] destination address The Default Address Selection for IPv6 [ADDRSEL] destination address
selection mechanism could save the application a few useless selection mechanism could save the application a few useless
connection attempts by placing the IPv4 addresses in front of the connection attempts by placing the IPv4 addresses in front of the
IPv6 addresses. This would be desired since all IPv6 destinations in IPv6 addresses. This would be desired since all IPv6 destinations in
this scenario are unreachable (there's no route to them), and the this scenario are unreachable (there's no route to them), and the
system's only IPv6 source address is inadequate to communicate with system's only IPv6 source address is inadequate to communicate with
off-link destinations even if it did have an off-link route. off-link destinations even if it did have an off-link route.
Let's examine how the destination address selection mechanism behaves Let's examine how the destination address selection mechanism behaves
in the face of this scenario when given one IPv4 destination and one in the face of this scenario when given one IPv4 destination and one
IPv6 destination. IPv6 destination.
The first rule, "Avoid unusable destinations", could prefer the IPv4 The first rule, "Avoid unusable destinations", would prefer the IPv4
destination over the IPv6 destination, but only if the IPv6 destination over the IPv6 destination, but only if the IPv6
destination is determined to be unreachable. The unreachability destination were determined to be unreachable. The unreachability
determination for a destination as it pertains to this rule is an determination for a destination as it pertains to this rule is an
implementation detail. One implementable method is to do a simple implementation detail. One implementable method is to do a simple
forwarding table lookup on the destination, and to deem the forwarding table lookup on the destination, and to deem the
destination as reachable if the lookup succeeds. The Neighbor destination as reachable if the lookup succeeds. The Neighbor
Discovery on-link assumption mentioned in Section 2.2 makes this Discovery on-link assumption mentioned in Section 2.2 makes this
method somewhat irrelevant, however, as an implementation of the method somewhat irrelevant, however, as an implementation of the
assumption could simply be to insert an IPv6 default on-link route assumption could simply be to insert an IPv6 default on-link route
into the system's forwarding table when the default router list is into the system's forwarding table when the default router list is
empty. The side-effect is that the rule would always determine that empty. The side-effect is that the rule would always determine that
all IPv6 destinations are reachable. Therefore, this rule will not all IPv6 destinations are reachable. Therefore, this rule will not
necessarily prefer one destination over the other. necessarily prefer one destination over the other.
The second rule, "Prefer matching scope", could prefer the IPv4 The second rule, "Prefer matching scope", could prefer the IPv4
destination over the IPv6 destination, but only if the IPv4 destination over the IPv6 destination, but only if the IPv4
destination's scope matches the scope of the system's IPv4 source destination's scope matches the scope of the system's IPv4 source
address. Since [ADDRSEL] considers private addresses (as defined in address. Since [ADDRSEL] considers private addresses (as defined in
[PRIVADDR]) of site-local scope, then this rule will not prefer [PRIVADDR]) of site-local scope, then this rule will not prefer
either destination over the other. The link-local IPv6 source either destination over the other. The link-local IPv6 source
doesn't match the global IPv6 destination, and the site-local IPv4 doesn't match the global IPv6 destination, and the "site-local" IPv4
source doesn't match the global IPv4 destination. The tie-breaking source doesn't match the global IPv4 destination. The tie-breaking
rule in this case is rule 6, "Prefer higher precedence". Since IPv6 rule in this case is rule 6, "Prefer higher precedence". Since IPv6
destinations are of higher precedence than IPv4 destinations in the destinations are of higher precedence than IPv4 destinations in the
default policy table, the IPv6 destination will be preferred. default policy table, the IPv6 destination will be preferred.
The solution in this case could be to add a new rule after rule 2 The solution in this case could be to add a new rule after rule 2
(rule 2.5) that avoids non-link-local IPv6 destinations whose source (rule 2.5) that avoids non-link-local IPv6 destinations whose
addresses are link-local. Of course, if the host is manually selected source addresses are link-local. Of course, if the host is
assigned a global IPv6 source address, then rule 2 will automatically manually assigned a global IPv6 source address, then rule 2 will
prefer the IPv6 destination, and there is no fix other than to make automatically prefer the IPv6 destination, and there is no fix other
sure rule 1 considers IPv6 destinations unreachable in this scenario. than to make sure rule 1 considers IPv6 destinations unreachable in
this scenario.
Fixing the destination address selection mechanism by adding such a Fixing the destination address selection mechanism by adding such a
rule is only a mitigating factor if applications use standard name rule is only a mitigating factor if applications use standard name
resolution API's that implement this mechanism, and these resolution API's that implement this mechanism, and these
applications try addresses in the order returned. This may not be an applications try addresses in the order returned. This may not be an
acceptable assumption in some cases, as there are applications that acceptable assumption in some cases, as there are applications that
use hard coded addresses and address search orders (DNS resolver is use hard coded addresses and address search orders and/or literal
one example), and/or literal addresses passed in from the user. Such addresses passed in from the user.
applications will obviously be subject to whatever connection delays
are associated with attempting a connection to an unreachable For example, one such application is the DNS resolver. In this case,
a configuration file usually contains a list of literal addresses to
be used as DNS name servers. The resolver client tries these servers
in the order that they appear in the file, bypassing address
selection rules.
Such applications will obviously be subject to whatever connection
delays are associated with attempting a connection to an unreachable
destination. This is discussed in more detail in the next few destination. This is discussed in more detail in the next few
sections. sections.
2.2 Neighbor Discovery's On-Link Assumption Considered Harmful 2.2 Neighbor Discovery's On-Link Assumption Considered Harmful
Let's assume that the application described in Section 2 is Let's assume that the application described in Section 2 is
attempting a connection to an IPv6 address first, either because the attempting a connection to an IPv6 address first, either because the
destination address selection mechanism described in Section 2.1 destination address selection mechanism described in Section 2.1
returned the addresses in that order, or because the application returned the addresses in that order, or because the application
isn't trying the addresses in the order returned. Regardless, the isn't trying the addresses in the order returned. Regardless, the
skipping to change at page 5, line 32 skipping to change at page 5, line 40
list is empty, then the host assumes that the destination is on-link. list is empty, then the host assumes that the destination is on-link.
This issue is described in detail in [ONLINK]. In summary, this This issue is described in detail in [ONLINK]. In summary, this
assumption makes the unreachability detection of off-link nodes in assumption makes the unreachability detection of off-link nodes in
the absence of a default router a lengthy operation. This is due to the absence of a default router a lengthy operation. This is due to
the cost of attempting Neighbor Discovery link-layer address the cost of attempting Neighbor Discovery link-layer address
resolution for each destination, and potential transport layer costs resolution for each destination, and potential transport layer costs
associated with connection timeouts. The transport layer issues are associated with connection timeouts. The transport layer issues are
discussed later in Section 2.3. discussed later in Section 2.3.
On a network that has no IPv6 routing and no IPv6 neighbors, making On a network that has no IPv6 routing and no IPv6 neighbors, making
the assumption that every IPv6 destination is on-link will be a the assumption that every IPv6 destination is on-link will be costly
costly and incorrect assumption. If an application has a list of and incorrect. If an application has a list of addresses associated
addresses associated with a destination and the first 15 are IPv6 with a destination and the first 15 are IPv6 addresses, then the
addresses, then the application won't be able to successfully send a application won't be able to successfully send a packet to the
packet to the destination until the attempts to resolve each IPv6 destination until the attempts to resolve each IPv6 address have
address have failed. This could take 45 seconds failed. This could take 45 seconds (MAX_MULTICAST_SOLICIT *
(MAX_MULTICAST_SOLICIT * RETRANS_TIMER * 15). This could be RETRANS_TIMER * 15). This could be compounded by any transport
compounded by any transport timeouts associated with each connection timeouts associated with each connection attempt, bringing the
attempt. timeouts to even dozens of minutes.
If IPv6 hosts don't assume that destinations are on-link as described If IPv6 hosts don't assume that destinations are on-link as described
above, then communication with destinations that are not on-link and above, then communication with destinations that are not on-link and
unreachable should immediately fail. The IPv6 implementation should unreachable should immediately fail. The IPv6 implementation should
be able to immediately notify applications or the transport layer be able to immediately notify applications or the transport layer
that it has no route to such IPv6 destinations, and applications that it has no route to such IPv6 destinations, so that applications
won't waste time waiting for address resolution to fail. won't waste time waiting for address resolution to fail.
If hosts need to communicate with on-link destinations in the absence If hosts need to communicate with on-link destinations in the absence
of default routers, then then they need to be explicitly configured of default routers, then then they need to be explicitly configured
to have on-link routes for those destinations. to have on-link routes for those destinations.
2.3 Transport Protocol Robustness 2.3 Transport Protocol Robustness
Making the same set of assumptions as Section 2.2, regardless of how Making the same set of assumptions as Section 2.2, regardless of how
long the network layer takes to determine that the IPv6 destination long the network layer takes to determine that the IPv6 destination
is unreachable, the delay associated with a connection attempt to an is unreachable, the delay associated with a connection attempt to an
unreachable destination can be compounded by the transport layer. unreachable destination can be compounded by the transport layer.
When the unreachability of a destination is obviated by the reception When the unreachability of a destination is obviated by the reception
of an ICMPv6 destination unreachable message, the transport layer of an ICMPv6 destination unreachable message, the transport layer
should make it possible for the application to deal with this by should make it possible for the application or API to deal with this.
failing the connection attempt, passing ICMPv6 errors up to the It could fail the connection attempt, pass ICMPv6 errors up to the
application, etc... Such messages would be received in the cases application, or pass them up to an API that is handling this for the
mentioned in Section 2 in which a node has no default routers and NUD application, etc.
fails for destinations assumed to be on-link, and when firewalls or
other systems that enforce scope boundaries send such ICMPv6 errors Such messages would be received in the cases mentioned in Section 2
as described in Section 3.1 and Section 3.3. in which a node has no default routers and NUD fails for destinations
assumed to be on-link, and when firewalls or other systems that
enforce scope boundaries send such ICMPv6 errors as described in
Section 3.1 and Section 3.3.
For cases when packets to a destination are essentially black-holed For cases when packets to a destination are essentially black-holed
and no ICMPv6 errors are generated, there is very little additional and no ICMPv6 errors are generated, there is very little additional
remedy other than the existing timer mechanisms inside transport remedy other than the existing timer mechanisms inside transport
layers and applications. The following transport layer implication layers and applications. The following transport layer implication
discussions deal with the former case, in which ICMPv6 errors are discussions deal with the former case, in which ICMPv6 errors are
received. received.
2.3.1 TCP Implications 2.3.1 TCP Implications
In the case of a socket application attempting a connection via TCP, In the case of a socket application attempting a connection via TCP,
it would be unreasonable for the application to block even after the it would be unreasonable for the application to block even after the
system has received notification that the destination address is system has received notification that the destination address is
unreachable via an ICMPv6 Destination Unreachable message. unreachable via an ICMPv6 Destination Unreachable message.
Following are some ways of solving TCP related delays associated with Following are some ways of solving TCP related delays associated with
destination unreachability when ICMPv6 errors are generated. destination unreachability when ICMPv6 errors are generated.
2.3.1.1 TCP Connection Termination 2.3.1.1 TCP Connection Termination
One solution is for TCP to abort connections in SYN-SENT or One solution for TCP is to abort connections in SYN-SENT or
SYN-RECEIVED state when it receives an ICMPv6 Destination Unreachable SYN-RECEIVED state when it receives an ICMPv6 Destination Unreachable
message. message.
It should be noted that the Requirements for Internet Hosts It should be noted that the Requirements for Internet Hosts
[HOSTREQS] document, in section 4.2.3.9., states that TCP MUST NOT [HOSTREQS] document, in section 4.2.3.9., states that TCP MUST NOT
abort connections when receiving ICMP Destination Unreachable abort connections when receiving ICMP Destination Unreachable
messages that indicate "soft errors", where soft errors are defined messages that indicate "soft errors", where soft errors are defined
as ICMP codes 0 (network unreachable), 1 (host unreachable), and 5 as ICMP codes 0 (network unreachable), 1 (host unreachable), and 5
(source route failed), and SHOULD abort connections upon receiving (source route failed), and SHOULD abort connections upon receiving
the other codes (which are considered "hard errors"). ICMPv6 didn't the other codes (which are considered "hard errors"). ICMPv6 didn't
skipping to change at page 7, line 17 skipping to change at page 7, line 28
When [HOSTREQS] was written, most applications would mostly only try When [HOSTREQS] was written, most applications would mostly only try
one address when establishing communication with a destination. Not one address when establishing communication with a destination. Not
aborting a connection was a sane thing to do if re-trying a single aborting a connection was a sane thing to do if re-trying a single
address was a better alternative over quitting the application address was a better alternative over quitting the application
altogether. With IPv6, and especially on dual stack systems, altogether. With IPv6, and especially on dual stack systems,
destinations are often assigned multiple addresses (at least one IPv4 destinations are often assigned multiple addresses (at least one IPv4
and one IPv6 address), and applications iterate through destination and one IPv6 address), and applications iterate through destination
addresses when attempting connections. addresses when attempting connections.
Since soft errors conditions are those that would entice an Since soft errors conditions are those that would cause an
application to continue iterating to another address, TCP shouldn't application to continue iterating to another address, TCP could
make the distinction between ICMPv6 soft errors and hard errors when simply not make the distinction between ICMPv6 soft errors and hard
in SYN-SENT or SYN-RECEIVED state. It should abort a connection in errors when in SYN-SENT or SYN-RECEIVED state. It would abort a
those states when receiving any ICMPv6 Destination Unreachable connection in those states when receiving any ICMPv6 Destination
message. When in any other state, TCP would behave as described in Unreachable message. When in any other state, TCP would behave as
[HOSTREQS]. described in [HOSTREQS].
Many TCP implementations already behave this way, but others do not. Many TCP implementations already behave this way, but others do not.
This should be noted as a best current practice in this case. This should be noted as a best current practice in this case.
A tangential method of handling the problem in this way would be for A tangential method of handling the problem in this way would be for
applications to somehow notify the TCP layer of their preference in applications to somehow notify the TCP layer of their preference in
the matter. An application could notify TCP that it should abort a the matter. An application could notify TCP that it should abort a
connection upon receipt of particular ICMPv6 errors. Similarly, it connection upon receipt of particular ICMPv6 errors. Similarly, it
could notify TCP that it should not abort a connection. This would could notify TCP that it should not abort a connection. This would
allow existing TCP implementations to maintain their status quo at allow existing TCP implementations to maintain their status quo at
the expense of increased application complexity. the expense of increased application complexity.
Drawbacks do exist for this TCP behavior. One drawback involves a
transient problem existing at the network layer that could cause all
destinations to be unreachable for a temporary length of time. In
such a situation, the application could quickly cycle through the
destinations and return an error, when it could have let TCP retry a
destination a few seconds later when the transient problem could have
been mitigated.
Another drawback is related to the insecurity of ICMP messages.
Reacting to "soft errors" in all TCP connection states would have
significant security implications, as it would be possible for anyone
to reset established sessions, even without IP spoofing. However,
this document only proposes reacting to such errors in SYN-SENT or
SYN-RECEIVED states. These states are very short-lived, so the
window of exposure is very short and the threat is negligible.
2.3.1.2 Asynchronous Application Notification 2.3.1.2 Asynchronous Application Notification
In section 4.2.4.1, [HOSTREQS] states that there MUST be a mechanism In section 4.2.4.1, [HOSTREQS] states that there MUST be a mechanism
for reporting soft TCP error conditions to the application. Such a for reporting soft TCP error conditions to the application. Such a
mechanism (assuming one is implemented) could be used by applications mechanism (assuming one is implemented) could be used by applications
to cycle through destination addresses. to cycle through destination addresses.
2.3.1.3 Higher Level API
Regardless of which solution is used, an API that handles connection
attempts to a destination in a transparent way could remove the
complexity from applications and handle this scenario gracefully. The
API could contain the intelligence required to resolve the hostname,
try each destination address, etc.
Such an API would also help to build applications that are protocol
independent, and would reduce the impact on applications when
transport layer behavior changes.
The drawback of this approach is the need to change applications to
use the API.
2.3.2 UDP Implications 2.3.2 UDP Implications
As noted in [HOSTREQS] section 4.1.3.3, UDP implementations MUST pass As noted in [HOSTREQS] section 4.1.3.3, UDP implementations MUST pass
to the application layer all ICMP error messages that it receives to the application layer all ICMP error messages that it receives
from the IP layer. As a result, proper handling destination from the IP layer, granted that the socket used is connected. As a
unreachability by UDP applications is the responsibility of the result, proper handling destination unreachability by UDP
applications themselves. applications is the responsibility of the applications themselves.
A higher level API as mentioned in Section 2.3.1.3 could remove the
complexity from applications in this case as well.
2.3.3 SCTP Implications 2.3.3 SCTP Implications
According to [SCTPIMP], SCTP ignores all ICMPv6 destination According to [SCTPIMP], SCTP ignores all ICMPv6 destination
unreachable messages. The existing SCTP specifications do not unreachable messages that do not indicate that the SCTP protocol
itself is unreachable. The existing SCTP specifications do not
suggest any action on the part of the implementation on reception of suggest any action on the part of the implementation on reception of
such messages. Investigation needs to be done to determine the a "soft error".
implications.
Unlike TCP, SCTP itself is multihomed and allows a user to specify a
list of addresses to connect to. Upon reception of a Destination
Unreachable Message during connection setup SCTP should attempt to
retransmit the Initiation message to one of its alternate addresses
and put the address that encountered the "soft error" into the
unreachable state. This will prevent use of the address after the
association is set-up until SCTP's heartbeat mechanism finds the
address to be reachable.
In some cases the application may not have provided a list of
addresses. In these cases it may be beneficial for SCTP, at a
minimum, to mark the address as unreachable so that the application
can acquire a notification through its application interface. In
such a case the application could then either take action upon the
address notification by aborting the connection attempt or allow
SCTP's normal timer mechanism to continue retransmitting the INIT
message to the peer's address.
There is work in progress to specify adding the support for handling
soft errors to the SCTP specification.
3. Other Problematic Scenarios 3. Other Problematic Scenarios
This section describes problems that could arise for a dual stack This section describes problems that could arise for a dual stack
system with IPv6 enabled when placed on a network with IPv6 system with IPv6 enabled when placed on a network with IPv6
connectivity. connectivity.
3.1 IPv6 Network of Smaller Scope 3.1 IPv6 Network of Smaller Scope
A network that has a smaller scope of connectivity for IPv6 as it A network that has a smaller scope of connectivity for IPv6 as it
skipping to change at page 9, line 39 skipping to change at page 11, line 9
Disabling IPv6 on the end nodes is another solution. The idea would Disabling IPv6 on the end nodes is another solution. The idea would
be that enabling IPv6 on dual stack nodes is a manual process that be that enabling IPv6 on dual stack nodes is a manual process that
would be done when the administrator knows that IPv6 connectivity is would be done when the administrator knows that IPv6 connectivity is
adequate. adequate.
3.3 Security 3.3 Security
Enabling IPv6 on a host implies that the services on the host may be Enabling IPv6 on a host implies that the services on the host may be
open to IPv6 communication. If the service itself is insecure and open to IPv6 communication. If the service itself is insecure and
depends on security policy enforced somewhere else on the network depends on a security policy enforced somewhere else on the network
(such as in a firewall), then there is potential for new attacks (such as in a firewall), then there is potential for new attacks
against the service. against the service.
A firewall, for example, may not be enforcing the same policy for A firewall may not be enforcing the same policy for IPv4 as for IPv6
IPv4 as for IPv6 traffic. One possibility is that the firewall could traffic, which could be due to misconfiguration of the firewall. One
have more relaxed policy for IPv6, perhaps by letting all IPv6 possibility is that the firewall could have more relaxed policy for
packets pass through, or by letting all IPv4 protocol 41 packets pass IPv6, perhaps by letting all IPv6 packets pass through, or by letting
through. In this scenario, the dual stack hosts within the protected all IPv4 protocol 41 packets pass through. In this scenario, the
network could be subject to different attacks than for IPv4. dual stack hosts within the protected network could be subject to
different attacks than for IPv4.
Even if a firewall has a stricter policy or identical policy for IPv6 Even if a firewall has a stricter policy or identical policy for IPv6
traffic than for IPv4 (the extreme case being that it drops all IPv6 traffic than for IPv4 (the extreme case being that it drops all IPv6
traffic), IPv6 packets could go through the network untouched if traffic), IPv6 packets could go through the network untouched if
tunneled over a transport layer. This could open the host to direct tunneled over a transport layer. This could open the host to direct
IPv6 attacks. IPv6 attacks. It should be noted that IPv4 packets can also be
tunneled, so this is not a new security concern for IPv6. Firewalls
must be deliberately and properly configured.
A similar problem could exist for VPN software. A VPN could protect A similar problem could exist for virtual private network (VPN)
all IPv4 packets but transmit all others onto the local subnet software. A VPN could protect all IPv4 packets but transmit all
unprotected. At least one widely used VPN behaves this way. This is others onto the local subnet unprotected. At least one widely used
problematic on a dual stack host that has IPv6 enabled on its local VPN behaves this way. This is problematic on a dual stack host that
network. It establishes its VPN link and attempts to communicate has IPv6 enabled on its local network. It establishes its VPN link
with destinations that resolve to both IPv4 and IPv6 addresses. The and attempts to communicate with destinations that resolve to both
destination address selection mechanism prefers the IPv6 destination IPv4 and IPv6 addresses. The destination address selection mechanism
so the application sends packets to an IPv6 address. The VPN doesn't prefers the IPv6 destination so the application sends packets to an
know about IPv6, so instead of protecting the packets and sending IPv6 address. The VPN doesn't know about IPv6, so instead of
them to the remote end of the VPN, it passes such packets in the protecting the packets and sending them to the remote end of the VPN,
clear to the local network. The reason that packets meant to be it passes such packets in the clear to the local network.
protected would be sent in the clear on the local network is either
because of the on-link assumption discussed in Section 2.2, or of This is problematic for a number of reasons. The first is that if
malicious hijacking of traffic by a rogue "fake" router advertising a the node has a default IPv6 route, the packets will be forwarded
prefix. off-link to an unknown destination. Another is if no legitimate
router is on-link and the node makes the on-link assumption discussed
in Section 2.2, the packets will simply be sent onto the local link
to be potentially viewed by a node spoofing the destination. A third
is if a rogue IPv6 router exists on-link. In that case the malicious
node will simply be sent all IPv6 packets in the clear.
3.3.1 Mitigating Security Risks 3.3.1 Mitigating Security Risks
The security policy implemented in firewalls, VPN software, or other The security policy implemented in firewalls, VPN software, or other
devices, must take a stance whether it applies equally to both IPv4 devices, must take a stance whether it applies equally to both IPv4
and IPv6 traffic. It is probably desirable for policy to apply and IPv6 traffic. It is probably desirable for the policy to apply
equally to both IPv4 and IPv6, but the most important thing is to be equally to both IPv4 and IPv6, but the most important thing is to be
aware of the potential problem, and to make the policy clear to the aware of the potential problem, and to make the policy clear to the
administrator and user. administrator and user.
There is still a risk that IPv6 packets could be tunneled over a There is still a risk that IPv6 packets could be tunneled over a
transport layer such as UDP, implicitly bypassing security policy. transport layer such as UDP, implicitly bypassing the security
Some more complex mechanism could be implemented to apply the correct policy. Some more complex mechanisms could be implemented to apply
policy to such packets. This could be easy to do if tunnel endpoints the correct policy to such packets. This could be easy to do if
are co-located with a firewall, but more difficult if internal nodes tunnel endpoints are co-located with a firewall, but more difficult
do their own IPv6 tunneling. if internal nodes do their own IPv6 tunneling.
4. Application Robustness 4. Application Robustness
Enabling IPv6 on a dual stack node is only useful if applications Enabling IPv6 on a dual stack node is only useful if applications
that support IPv6 on that node properly cycle through addresses that support IPv6 on that node properly cycle through addresses
returned from name lookups and fall back to IPv4 when IPv6 returned from name lookups and fall back to IPv4 when IPv6
communication fails. Simply cycling through the list of addresses communication fails. Simply cycling through the list of addresses
returned from a name lookup when attempting connections works in most returned from a name lookup when attempting connections works in most
cases for most applications, but there are still cases where that's cases for most applications, but there are still cases where that's
not enough. Applications also need to be aware that the fact that a not enough. Applications also need to be aware that the fact that a
dual stack destination's IPv6 address is published in the DNS does dual stack destination's IPv6 address is published in the DNS does
not necessarily imply that all services on that destination function not necessarily imply that all services on that destination function
over IPv6. This problem, along with a thorough discussion of IPv6 over IPv6. This problem, along with a thorough discussion of IPv6
application transition guidelines, is discussed in [APPTRANS]. application transition guidelines, is discussed in [APPTRANS].
5. Security Considerations 5. Security Considerations
This document raises security concerns in Section 3.3. This document raises security concerns in Section 3.3. They are
summarized below:
Normative References o Firewalls need to be configured properly to have deliberate
security policies for IPv6 packets, including IPv6 packets
encapsulated in other layers.
o Implementations of virtual private networks need to have a
deliberate IPv6 security policy that doesn't allow packets to
accidentally appear in the clear when they were intended to be
sent securely over the VPN.
6. References
6.1 Normative References
[ADDRSEL] Draves, R., "Default Address Selection for Internet [ADDRSEL] Draves, R., "Default Address Selection for Internet
Protocol version 6 (IPv6)", RFC 3484, February 2003. Protocol version 6 (IPv6)", RFC 3484, February 2003.
[APPTRANS] [APPTRANS]
Hong, Y-G., Hagino, J., Savola, P. and M. Castro, Hong, Y-G., Hagino, J., Savola, P. and M. Castro,
"Application Aspects of IPv6 Transition", October 2003. "Application Aspects of IPv6 Transition", March 2004.
draft-ietf-v6ops-application-transition-00 draft-ietf-v6ops-application-transition-02
[ND] Narten, T., Nordmark, E. and W. Simpson, "Neighbor [ND] Narten, T., Nordmark, E. and W. Simpson, "Neighbor
Discovery for IP Version 6 (IPv6)", RFC 2461, December Discovery for IP Version 6 (IPv6)", RFC 2461, December
1998. 1998.
[ONLINK] Roy, S., Durand, A. and J. Paugh, "IPv6 Neighbor Discovery [ONLINK] Roy, S., Durand, A. and J. Paugh, "IPv6 Neighbor Discovery
On-Link Assumption Considered Harmful", October 2003. On-Link Assumption Considered Harmful", October 2003.
draft-ietf-v6ops-onlinkassumption-00 draft-ietf-v6ops-onlinkassumption-02
Informative References 6.2 Informative References
[AUTOCONF] [AUTOCONF]
Thomson, S. and T. Narten, "IPv6 Stateless Address Thomson, S. and T. Narten, "IPv6 Stateless Address
Autoconfiguration", RFC 2462, December 1998. Autoconfiguration", RFC 2462, December 1998.
[HOSTREQS] [HOSTREQS]
Braden, R., "Requirements for Internet Hosts -- Braden, R., "Requirements for Internet Hosts --
Communication Layers", STD 3, RFC 1122, October 1989. Communication Layers", STD 3, RFC 1122, October 1989.
[PRIVADDR] [PRIVADDR]
skipping to change at page 12, line 34 skipping to change at page 14, line 23
Sun Microsystems, Inc. Sun Microsystems, Inc.
17 Network Circle 17 Network Circle
UMPK17-202 UMPK17-202
Menlo Park, CA 94025 Menlo Park, CA 94025
EMail: james.paugh@sun.com EMail: james.paugh@sun.com
Appendix A. Acknowledgments Appendix A. Acknowledgments
The authors gratefully acknowledge the contributions of Jim Bound, The authors gratefully acknowledge the contributions of Jim Bound,
Tim Hartrick, Mika Liljeberg, Erik Nordmark, Pekka Savola, and Ronald Fernando Gont, Tony Hain, Tim Hartrick, Mika Liljeberg, Erik
Nordmark, Kacheong Poon, Pekka Savola, Randall Stewart, and Ronald
van der Pol. van der Pol.
Appendix B. Changes from draft-ietf-v6ops-v6onbydefault-00 Appendix B. Changes from draft-ietf-v6ops-v6onbydefault-01
o Added specificity to the DNS resolver problem in Section 2.1.
o Added a few paragraphs in Section 2.3.1.1 describing potential
drawbacks to TCP aborting connections ins SYN-SENT or SYN-RECEIVED
state.
o Added Section 2.3.1.3 describing how a higher level API could be
used to manage connections.
o Expanded Section 2.3.3 to describe desired SCTP behavior when
encountering soft errors.
o Added a summary of security concerns to Section 5.
o Miscellaneous editorial changes.
Appendix C. Changes from draft-ietf-v6ops-v6onbydefault-00
o Clarified in the abstract and introduction that the document is o Clarified in the abstract and introduction that the document is
meant to raise awareness, and not to specify whether IPv6 should meant to raise awareness, and not to specify whether IPv6 should
be enabled by default or not. be enabled by default or not.
o Shortened section Section 2.2 and made reference to [ONLINK]. o Shortened Section 2.2 and made reference to [ONLINK].
o Added clarification in section Section 2.3 about packets that are o Added clarification in Section 2.3 about packets that are lost
lost without ICMPv6 notification. without ICMPv6 notification.
o Section Section 2.3 now has subsections for TCP, UDP, and SCTP. o Section 2.3 now has subsections for TCP, UDP, and SCTP.
o Removed text in Section 2.3.1.1 suggesting that hosts usually were o Removed text in Section 2.3.1.1 suggesting that hosts usually were
only assigned one address when [HOSTREQS] was written. only assigned one address when [HOSTREQS] was written.
o Added text in Section 2.3.1.1 suggesting a method for applications o Added text in Section 2.3.1.1 suggesting a method for applications
to advise TCP of their preference for ICMPv6 handling. to advise TCP of their preference for ICMPv6 handling.
o Added section Section 2.3.1.2. o Added Section 2.3.1.2.
o Added section Section 2.3.2. o Added Section 2.3.2.
o Added section Section 2.3.3. o Added Section 2.3.3.
o Strengthened wording in section Section 3.1.1 to suggest that o Strengthened wording in Section 3.1.1 to suggest that devices
devices enforcing scope boundaries must send ICMPv6 Destination enforcing scope boundaries must send ICMPv6 Destination
Unreachable messages. Unreachable messages.
o Clarified that the VPN problem described in Section 3.3 is due to o Clarified that the VPN problem described in Section 3.3 is due to
a combination of the VPN software and either the on-link a combination of the VPN software and either the on-link
assumption and/or a "bad guy". assumption and/or a "bad guy".
o Shortened section Section 4 and made reference to [APPTRANS]. o Shortened Section 4 and made reference to [APPTRANS].
o Miscellaneous editorial changes. o Miscellaneous editorial changes.
Intellectual Property Statement Intellectual Property Statement
The IETF takes no position regarding the validity or scope of any The IETF takes no position regarding the validity or scope of any
intellectual property or other rights that might be claimed to intellectual property or other rights that might be claimed to
pertain to the implementation or use of the technology described in pertain to the implementation or use of the technology described in
this document or the extent to which any license under such rights this document or the extent to which any license under such rights
might or might not be available; neither does it represent that it might or might not be available; neither does it represent that it
 End of changes. 

This html diff was produced by rfcdiff 1.23, available from http://www.levkowetz.com/ietf/tools/rfcdiff/