draft-ietf-ippm-tcp-throughput-tm-05.txt   draft-ietf-ippm-tcp-throughput-tm-06.txt 
Network Working Group B. Constantine Network Working Group B. Constantine
Internet-Draft JDSU Internet-Draft JDSU
Intended status: Informational G. Forget Intended status: Informational G. Forget
Expires: February 12, 2011 Bell Canada (Ext. Consultant) Expires: February 27, 2011 Bell Canada (Ext. Consultant)
L. Jorgenson L. Jorgenson
nooCore nooCore
Reinhard Schrage Reinhard Schrage
Schrage Consulting Schrage Consulting
August 12, 2010 August 27, 2010
TCP Throughput Testing Methodology TCP Throughput Testing Methodology
draft-ietf-ippm-tcp-throughput-tm-05.txt draft-ietf-ippm-tcp-throughput-tm-06.txt
Abstract Abstract
This memo describes a methodology for measuring sustained TCP This memo describes a methodology for measuring sustained TCP
throughput performance in an end-to-end managed network environment. throughput performance in an end-to-end managed network environment.
This memo is intended to provide a practical approach to help users This memo is intended to provide a practical approach to help users
validate the TCP layer performance of a managed network, which should validate the TCP layer performance of a managed network, which should
provide a better indication of end-user application level experience. provide a better indication of end-user application level experience.
In the methodology, various TCP and network parameters are identified In the methodology, various TCP and network parameters are identified
that should be tested as part of the network verification at the TCP that should be tested as part of the network verification at the TCP
layer. layer.
Requirements Language
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC 2119 [RFC2119].
Status of this Memo Status of this Memo
This Internet-Draft is submitted to IETF in full conformance with the This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79. provisions of BCP 78 and BCP 79.
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 Task Force (IETF). Note that other groups may also distribute
other groups may also distribute working documents as Internet- working documents as Internet-Drafts. The list of current Internet-
Drafts. Creation date August 12, 2010. Drafts is at http://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."
The list of current Internet-Drafts can be accessed at This Internet-Draft will expire on February 27, 2011.
http://www.ietf.org/ietf/1id-abstracts.txt.
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html.
This Internet-Draft will expire on February 12, 2011.
Copyright Notice Copyright Notice
Copyright (c) 2010 IETF Trust and the persons identified as the Copyright (c) 2010 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
(http://trustee.ietf.org/license-info) in effect on the date of (http://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 BSD License. described in the Simplified BSD License.
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Goals of this Methodology. . . . . . . . . . . . . . . . . . . 4 2. Goals of this Methodology. . . . . . . . . . . . . . . . . . . 4
2.1 TCP Equilibrium State Throughput . . . . . . . . . . . . . 5 2.1 TCP Equilibrium State Throughput . . . . . . . . . . . . . 5
2.2 Metrics for TCP Throughput Tests . . . . . . . . . . . . . 6 2.2 Metrics for TCP Throughput Tests . . . . . . . . . . . . . 6
3. TCP Throughput Testing Methodology . . . . . . . . . . . . . . 7 3. TCP Throughput Testing Methodology . . . . . . . . . . . . . . 7
3.1 Determine Network Path MTU . . . . . . . . . . . . . . . . 8 3.1 Determine Network Path MTU . . . . . . . . . . . . . . . . 8
3.2. Baseline Round-trip Delay and Bandwidth. . . . . . . . . . 10 3.2. Baseline Round-trip Delay and Bandwidth. . . . . . . . . . 10
skipping to change at page 2, line 41 skipping to change at page 2, line 41
3.3. TCP Throughput Tests . . . . . . . . . . . . . . . . . . . 11 3.3. TCP Throughput Tests . . . . . . . . . . . . . . . . . . . 11
3.3.1 Calculate Optimum TCP Window Size. . . . . . . . . . . 12 3.3.1 Calculate Optimum TCP Window Size. . . . . . . . . . . 12
3.3.2 Conducting the TCP Throughput Tests. . . . . . . . . . 14 3.3.2 Conducting the TCP Throughput Tests. . . . . . . . . . 14
3.3.3 Single vs. Multiple TCP Connection Testing . . . . . . 15 3.3.3 Single vs. Multiple TCP Connection Testing . . . . . . 15
3.3.4 Interpretation of the TCP Throughput Results . . . . . 16 3.3.4 Interpretation of the TCP Throughput Results . . . . . 16
3.4. Traffic Management Tests . . . . . . . . . . . . . . . . . 16 3.4. Traffic Management Tests . . . . . . . . . . . . . . . . . 16
3.4.1 Traffic Shaping Tests. . . . . . . . . . . . . . . . . 16 3.4.1 Traffic Shaping Tests. . . . . . . . . . . . . . . . . 16
3.4.1.1 Interpretation of Traffic Shaping Test Results. . . 17 3.4.1.1 Interpretation of Traffic Shaping Test Results. . . 17
3.4.2 RED Tests. . . . . . . . . . . . . . . . . . . . . . . 18 3.4.2 RED Tests. . . . . . . . . . . . . . . . . . . . . . . 18
3.4.2.1 Interpretation of RED Results . . . . . . . . . . . 18 3.4.2.1 Interpretation of RED Results . . . . . . . . . . . 18
4. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 19 4. Security Considerations . . . . . . . . . . . . . . . . . . . 18
5. References . . . . . . . . . . . . . . . . . . . . . . . . . . 19 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 19
5.1. Registry Specification . . . . . . . . . . . . . . . . . . 19
5.2. Registry Contents . . . . . . . . . . . . . . . . . . . . 19
6. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 19
7. References . . . . . . . . . . . . . . . . . . . . . . . . . . 19
7.1 Normative References . . . . . . . . . . . . . . . . . . . 19
7.2 Informative References . . . . . . . . . . . . . . . . . . 20
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 20 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 20
1. Introduction 1. Introduction
Testing an operational network prior to customer activation is referred Testing an operational network prior to customer activation is
to as "turn-up" testing and the SLA is generally Layer 2/3 packet referred to as "turn-up" testing and the SLA (Service Level
throughput, delay, loss and jitter. Agreement) is generally based upon Layer 2/3 packet throughput,
delay, loss and jitter.
Network providers are coming to the realization that Layer 2/3 testing Network providers are coming to the realization that Layer 2/3
and TCP layer testing are required to more adequately ensure end-user testing and TCP layer testing are required to more adequately ensure
satisfaction. Therefore, the network provider community desires to end-user satisfaction. Therefore, the network provider community
measure network throughput performance at the TCP layer. Measuring desires to measure network throughput performance at the TCP layer.
TCP throughput provides a meaningful measure with respect to the end Measuring TCP throughput provides a meaningful measure with respect
user's application SLA (and ultimately reach some level of TCP to the end user experience (and ultimately reach some level of
testing interoperability which does not exist today). TCP testing interoperability which does not exist today).
Additionally, end-users (business enterprises) seek to conduct Additionally, end-users (business enterprises) seek to conduct
repeatable TCP throughput tests between enterprise locations. Since repeatable TCP throughput tests between enterprise locations. Since
these enterprises rely on the networks of the providers, a common test these enterprises rely on the networks of the providers, a common
methodology (and metrics) would be equally beneficial to both parties. test methodology (and metrics) would be equally beneficial to both
parties.
So the intent behind this TCP throughput draft is to define So the intent behind this TCP throughput draft is to define
a methodology for testing sustained TCP layer performance. In this a methodology for testing sustained TCP layer performance. In this
document, sustained TCP throughput is that amount of data per unit document, sustained TCP throughput is that amount of data per unit
time that TCP transports during equilibrium (steady state), i.e. time that TCP transports during equilibrium (steady state), i.e.
after the initial slow start phase. We refer to this state as TCP after the initial slow start phase. We refer to this state as TCP
Equilibrium, and that the equalibrium throughput is the maximum Equilibrium, and that the equilibrium throughput is the maximum
achievable for the TCP connection(s). achievable for the TCP connection(s).
There are many variables to consider when conducting a TCP throughput There are many variables to consider when conducting a TCP throughput
test and this methodology focuses on some of the most common test and this methodology focuses on some of the most common
parameters that should be considered such as: parameters that should be considered such as:
- Path MTU and Maximum Segment Size (MSS) - Path MTU and Maximum Segment Size (MSS)
- RTT and Bottleneck BW - RTT and Bottleneck BW
- Ideal TCP Window (Bandwidth Delay Product) - Ideal TCP Window (Bandwidth Delay Product)
- Single Connection and Multiple Connection testing - Single Connection and Multiple Connection testing
One other important note, it is highly recommended that traditional One other important note, it is highly recommended that traditional
Layer 2/3 type tests are conducted to verify the integrity of the Layer 2/3 type tests are conducted to verify the integrity of the
network before conducting TCP tests. Examples include RFC2544, iperf network before conducting TCP tests. Examples include RFC 2544
(UDP mode), or manual packet layer test techniques where packet [RFC2544], iperf (UDP mode), or manual packet layer test techniques
throughput, loss, and delay measurements are conducted. where packet throughput, loss, and delay measurements are conducted.
2. Goals of this Methodology 2. Goals of this Methodology
Before defining the goals of this methodology, it is important to Before defining the goals of this methodology, it is important to
clearly define the areas that are not intended to be measured or clearly define the areas that are not intended to be measured or
analyzed by such a methodology. analyzed by such a methodology.
- The methodology is not intended to predict TCP throughput - The methodology is not intended to predict TCP throughput
behavior during the transient stages of a TCP connection, such behavior during the transient stages of a TCP connection, such
as initial slow start. as initial slow start.
skipping to change at page 4, line 43 skipping to change at page 4, line 43
- Provide a practical test approach that specifies the more well - Provide a practical test approach that specifies the more well
understood (and end-user configurable) TCP parameters such as Window understood (and end-user configurable) TCP parameters such as Window
size, MSS (Maximum Segment Size), # connections, and how these affect size, MSS (Maximum Segment Size), # connections, and how these affect
the outcome of TCP performance over a network. the outcome of TCP performance over a network.
- Provide specific test conditions (link speed, RTT, window size, - Provide specific test conditions (link speed, RTT, window size,
etc.) and maximum achievable TCP throughput under TCP Equilbrium etc.) and maximum achievable TCP throughput under TCP Equilbrium
conditions. For guideline purposes, provide examples of these test conditions. For guideline purposes, provide examples of these test
conditions and the maximum achievable TCP throughput during the conditions and the maximum achievable TCP throughput during the
equilbrium state. Section 2.1 provides specific details concerning equilibrium state. Section 2.1 provides specific details concerning
the definition of TCP Equilibrium within the context of this draft. the definition of TCP Equilibrium within the context of this draft.
- Define two (2) basic metrics that can be used to compare the - Define two (2) basic metrics that can be used to compare the
performance of TCP connections under various network conditions performance of TCP connections under various network conditions
- In test situations where the recommended procedure does not yield - In test situations where the recommended procedure does not yield
the maximum achievable TCP throughput result, this draft provides some the maximum achievable TCP throughput result, this draft provides
possible areas within the end host or network that should be some possible areas within the end host or network that should be
considered for investigation (although again, this draft is not considered for investigation (although again, this draft is not
intended to provide a detailed diagnosis of these issues) intended to provide a detailed diagnosis of these issues)
2.1 TCP Equilibrium State Throughput 2.1 TCP Equilibrium State Throughput
TCP connections have three (3) fundamental congestion window phases TCP connections have three (3) fundamental congestion window phases
as documented in RFC2581. These states are: as documented in RFC 5681 [RFC5681]. These states are:
- Slow Start, which occurs during the beginning of a TCP transmission - Slow Start, which occurs during the beginning of a TCP transmission
or after a retransmission time out event or after a retransmission time out event
- Congestion avoidance, which is the phase during which TCP ramps up - Congestion avoidance, which is the phase during which TCP ramps up
to establish the maximum attainable throughput on an end-end network to establish the maximum attainable throughput on an end-end network
path. Retransmissions are a natural by-product of the TCP congestion path. Retransmissions are a natural by-product of the TCP congestion
avoidance algorithm as it seeks to achieve maximum throughput on avoidance algorithm as it seeks to achieve maximum throughput on
the network path. the network path.
- Retransmission phase, which include Fast Retransmit (Tahoe) and Fast - Retransmission phase, which include Fast Retransmit (Tahoe) and
Recovery (Reno and New Reno). When a packet is lost, the Congestion Fast Recovery (Reno and New Reno). When a packet is lost, the
avoidance phase transitions to a Fast Retransmission or Recovery Congestion avoidance phase transitions to a Fast Retransmission or
Phase dependent upon the TCP implementation. Recovery Phase dependent upon the TCP implementation.
The following diagram depicts these states. The following diagram depicts these states.
| ssthresh | ssthresh
TCP | | TCP | |
Through- | | Equilibrium Through- | | Equilibrium
put | |\ /\/\/\/\/\ Retransmit /\/\ ... put | |\ /\/\/\/\/\ Retransmit /\/\ ...
| | \ / | Time-out / | | \ / | Time-out /
| | \ / | _______ _/ | | \ / | _______ _/
| Slow _/ |/ | / | Slow _/ | Slow _/ |/ | / | Slow _/
| Start _/ Congestion |/ |Start_/ Congestion | Start _/ Congestion |/ |Start_/ Congestion
| _/ Avoidance Loss | _/ Avoidance | _/ Avoidance Loss | _/ Avoidance
| _/ Event | _/ | _/ Event | _/
| _/ |/ | _/ |/
|/___________________________________________________________ |/__________________________________________________________
Time Time
This TCP methodology provides guidelines to measure the equilibrium This TCP methodology provides guidelines to measure the equilibrium
throughput which refers to the maximum sustained rate obtained by throughput which refers to the maximum sustained rate obtained by
congestion avoidance before packet loss conditions occur (which would congestion avoidance before packet loss conditions occur (which would
cause the state change from congestion avoidance to a retransmission cause the state change from congestion avoidance to a retransmission
phase). All maximum achievable throughputs specified in Section 3 are phase). All maximum achievable throughputs specified in Section 3 are
with respect to this Equilibrium state. with respect to this equilibrium state.
2.2 Metrics for TCP Throughput Tests 2.2 Metrics for TCP Throughput Tests
This draft focuses on a TCP throughtput methodology and also This draft focuses on a TCP throughput methodology and also
provides two basic metrics to compare results of various throughput provides two basic metrics to compare results of various throughput
tests. It is recognized that the complexity and unpredictability of tests. It is recognized that the complexity and unpredictability of
TCP makes it impossible to develop a complete set of metrics that TCP makes it impossible to develop a complete set of metrics that
account for the myriad of variables (i.e. RTT variation, loss account for the myriad of variables (i.e. RTT variation, loss
conditions, TCP implementation, etc.). However, these two basic conditions, TCP implementation, etc.). However, these two basic
metrics faciliate TCP throughput comparisons under varying network metrics will faciliate TCP throughput comparisons under varying
conditions and between network traffic management techniques. network conditions and between network traffic management techniques.
The TCP Efficiency metric is the percentage of bytes that were not The TCP Efficiency metric is the percentage of bytes that were not
retransmitted and is defined as: retransmitted and is defined as:
Transmitted Bytes - Retransmitted Bytes Transmitted Bytes - Retransmitted Bytes
--------------------------------------- x 100 --------------------------------------- x 100
Transmitted Bytes Transmitted Bytes
This metric provides a comparative measure between various QoS This metric provides a comparative measure between various QoS
mechanisms such as traffic management, congestion avoidance, and also mechanisms such as traffic management, congestion avoidance, and also
various TCP implementations (i.e. Reno, Vegas, etc.). various TCP implementations (i.e. Reno, Vegas, etc.).
As an example, if 100,000 bytes were sent and 2,000 had to be As an example, if 100,000 bytes were sent and 2,000 had to be
retransmitted, the TCP Efficiency would be calculated as: retransmitted, the TCP Efficiency would be calculated as:
100,000 - 2,000 100,000 - 2,000
---------------- x 100 = 98% ---------------- x 100 = 98%
100,000 100,000
Note that the retranmitted bytes may have occurred more than once, Note that the retransmitted bytes may have occurred more than once,
and these multiple retransmissions are added to the bytes retransmitted and these multiple retransmissions are added to the bytes
count. retransmitted count.
The second metric is the TCP Transfer Time, which is simply the time The second metric is the TCP Transfer Time, which is simply the time
it takes to transfer a block of data across simultaneous TCP it takes to transfer a block of data across simultaneous TCP
connections. The concept is useful when benchmarking traffic connections. The concept is useful when benchmarking traffic
management techniques, where multiple connections are generally management techniques, where multiple connections are generally
required. required.
The TCP Transfer time can also be used to provide a normalized ratio of The TCP Transfer time can also be used to provide a normalized ratio
the actual TCP Transfer Time versus ideal Transfer Time. This ratio of the actual TCP Transfer Time versus ideal Transfer Time. This
is called the TCP Transfer Index and is defined as: ratio is called the TCP Transfer Index and is defined as:
Actual TCP Transfer Time Actual TCP Transfer Time
------------------------- -------------------------
Ideal TCP Transfer Time Ideal TCP Transfer Time
An example would be the bulk transfer of 100 MB upon 5 simultaneous TCP An example would be the bulk transfer of 100 MB upon 5 simultaneous
connections over a 500 Mbit/s Ethernet service (each connection TCP connections over a 500 Mbit/s Ethernet service (each connection
uploading 100 MB). Each connection may achieve different throughputs uploading 100 MB). Each connection may achieve different throughputs
during a test and the overall throughput rate is not always easy to during a test and the overall throughput rate is not always easy to
determine (especially as the number of connections increases). determine (especially as the number of connections increases).
The ideal TCP Transfer Time would be ~8 seconds, but in this example, The ideal TCP Transfer Time would be ~8 seconds, but in this example,
the actual TCP Transfer Time was 12 seconds. The TCP Transfer Index the actual TCP Transfer Time was 12 seconds. The TCP Transfer Index
would be 12/8 = 1.5, which indicates that the transfer across all would be 12/8 = 1.5, which indicates that the transfer across all
connections took 1.5 times longer than the ideal. connections took 1.5 times longer than the ideal.
Note that both the TCP Efficiency and TCP Transfer Time metrics must be Note that both the TCP Efficiency and TCP Transfer Time metrics must
measured during each throughput test. The correlation of TCP Transfer be measured during each throughput test. The correlation of TCP
Time with TCP Efficiency can help to diagnose whether the TCP Transfer Transfer Time with TCP Efficiency can help to diagnose whether the
Time was negatively impacted by retransmissions (poor TCP Efficiency). TCP Transfer Time was negatively impacted by retransmissions (poor
TCP Efficiency).
3. TCP Throughput Testing Methodology 3. TCP Throughput Testing Methodology
As stated in Section 1, it is considered best practice to verify As stated in Section 1, it is considered best practice to verify
the integrity of the network by conducting Layer2/3 stress tests the integrity of the network by conducting Layer2/3 stress tests
such as RFC2544 (or other methods of network stress tests). If the such as RFC2544 (or other methods of network stress tests). If the
network is not performing properly in terms of packet loss, jitter, network is not performing properly in terms of packet loss, jitter,
etc. then the TCP layer testing will not be meaningful since the etc. then the TCP layer testing will not be meaningful since the
equalibrium throughput would be very difficult to achieve (in a equilibrium throughput would be very difficult to achieve (in a
"dysfunctional" network). "dysfunctional" network).
The following represents the sequential order of steps to conduct the The following represents the sequential order of steps to conduct the
TCP throughput testing methodology: TCP throughput testing methodology:
1. Identify the Path MTU. Packetization Layer Path MTU Discovery 1. Identify the Path MTU. Packetization Layer Path MTU Discovery
or PLPMTUD (RFC4821) should be conducted to verify the minimum network or PLPMTUD, [RFC4821], should be conducted to verify the maximum
path MTU. Conducting PLPMTUD establishes the upper limit for the MSS network path MTU. Conducting PLPMTUD establishes the upper limit for
to be used in subsequent steps. the MSS to be used in subsequent steps.
2. Baseline Round-trip Delay and Bandwidth. These measurements provide 2. Baseline Round-trip Delay and Bandwidth. These measurements
estimates of the ideal TCP window size, which will be used in provide estimates of the ideal TCP window size, which will be used in
subsequent test steps. subsequent test steps.
3. TCP Connection Throughput Tests. With baseline measurements 3. TCP Connection Throughput Tests. With baseline measurements
of round trip delay and bandwidth, a series of single and multiple TCP of round trip delay and bandwidth, a series of single and multiple
connection throughput tests can be conducted to baseline the network TCP connection throughput tests can be conducted to baseline the
performance expectations. network performance expectations.
4. Traffic Management Tests. Various traffic management and queuing 4. Traffic Management Tests. Various traffic management and queueing
techniques are tested in this step, using multiple TCP connections. techniques are tested in this step, using multiple TCP connections.
Multiple connection testing can verify that the network is configured Multiple connection testing can verify that the network is configured
properly for traffic shaping versus policing, various queuing properly for traffic shaping versus policing, various queueing
implementations, and RED. implementations, and RED.
Important to note are some of the key characteristics and Important to note are some of the key characteristics and
considerations for the TCP test instrument. The test host may be a considerations for the TCP test instrument. The test host may be a
standard computer or dedicated communications test instrument standard computer or dedicated communications test instrument
and these TCP test hosts be capable of emulating both a client and a and these TCP test hosts be capable of emulating both a client and a
server. server.
Whether the TCP test host is a standard computer or dedicated test Whether the TCP test host is a standard computer or dedicated test
instrument, the following areas should be considered when selecting instrument, the following areas should be considered when selecting
a test host: a test host:
- TCP implementation used by the test host OS, i.e. Linux OS kernel - TCP implementation used by the test host OS, i.e. Linux OS kernel
using TCP Reno, TCP options supported, etc. This will obviously be using TCP Reno, TCP options supported, etc. This will obviously be
more important when using custom test equipment where the TCP more important when using custom test equipment where the TCP
implementation may be customized or tuned to run in higher implementation may be customized or tuned to run in higher
performance hardware performance hardware
- Most importantly, the TCP test host must be capable of generating - Most importantly, the TCP test host must be capable of generating
and receiving stateful TCP test traffic at the full link speed of the and receiving stateful TCP test traffic at the full link speed of the
network under test. As a general rule of thumb, testing TCP throughput network under test. As a general rule of thumb, testing TCP
at rates greater than 100 Mbit/sec generally requires high throughput at rates greater than 100 Mbit/sec generally requires high
performance server hardware or dedicated hardware based test tools. performance server hardware or dedicated hardware based test tools.
- To measure RTT and TCP Efficiency per connection, this will generally - To measure RTT and TCP Efficiency per connection, this will
require dedicated hardware based test tools. In the absence of generally require dedicated hardware based test tools. In the absence
dedciated hardware based test tools, these measurements may need to be of dedicated hardware based test tools, these measurements may need
conducted with packet capture tools (conduct TCP throughput tests and to be conducted with packet capture tools (conduct TCP throughput
analyze RTT and retransmission results with packet captures). tests and analyze RTT and retransmission results with packet
captures).
3.1. Determine Network Path MTU 3.1. Determine Network Path MTU
TCP implementations should use Path MTU Discovery techniques (PMTUD). TCP implementations should use Path MTU Discovery techniques (PMTUD).
PMTUD relies on ICMP 'need to frag' messages to learn the path MTU. PMTUD relies on ICMP 'need to frag' messages to learn the path MTU.
When a device has a packet to send which has the Don't Fragment (DF) When a device has a packet to send which has the Don't Fragment (DF)
bit in the IP header set and the packet is larger than the Maximum bit in the IP header set and the packet is larger than the Maximum
Transmission Unit (MTU) of the next hop link, the packet is dropped Transmission Unit (MTU) of the next hop link, the packet is dropped
and the device sends an ICMP 'need to frag' message back to the host and the device sends an ICMP 'need to frag' message back to the host
that originated the packet. The ICMP 'need to frag' message includes that originated the packet. The ICMP 'need to frag' message includes
skipping to change at page 10, line 16 skipping to change at page 10, line 16
Before stateful TCP testing can begin, it is important to baseline Before stateful TCP testing can begin, it is important to baseline
the round trip delay and bandwidth of the network to be tested. the round trip delay and bandwidth of the network to be tested.
These measurements provide estimates of the ideal TCP window size, These measurements provide estimates of the ideal TCP window size,
which will be used in subsequent test steps. These latency and which will be used in subsequent test steps. These latency and
bandwidth tests should be run during the time of day for which bandwidth tests should be run during the time of day for which
the TCP throughput tests will occur. the TCP throughput tests will occur.
The baseline RTT is used to predict the bandwidth delay product and The baseline RTT is used to predict the bandwidth delay product and
the TCP Transfer Time for the subsequent throughput tests. Since this the TCP Transfer Time for the subsequent throughput tests. Since this
methodology requires that RTT be measured during the entire throughput methodology requires that RTT be measured during the entire
test, the extent by which the RTT varied during the throughput test can throughput test, the extent by which the RTT varied during the
be quantified. throughput test can be quantified.
3.2.1 Techniques to Measure Round Trip Time 3.2.1 Techniques to Measure Round Trip Time
Following the definitions used in the references of the appendix; Following the definitions used in the references of the appendix;
Round Trip Time (RTT) is the time elapsed between the clocking in of Round Trip Time (RTT) is the time elapsed between the clocking in of
the first bit of a payload packet to the receipt of the last bit of the the first bit of a payload packet to the receipt of the last bit of
corresponding acknowledgement. Round Trip Delay (RTD) is used the corresponding acknowledgement. Round Trip Delay (RTD) is used
synonymously to twice the Link Latency. synonymously to twice the Link Latency.
In any method used to baseline round trip delay between network In any method used to baseline round trip delay between network
end-points, it is important to realize that network latency is the end-points, it is important to realize that network latency is the
sum of inherent network delay and congestion. The RTT should be sum of inherent network delay and congestion. The RTT should be
baselined during "off-peak" hours to obtain a reliable figure for baselined during "off-peak" hours to obtain a reliable figure for
network latency (versus additional delay caused by congestion). network latency (versus additional delay caused by congestion).
During the actual sustained TCP throughput tests, it is critical During the actual sustained TCP throughput tests, it is critical
to measure RTT along with measured TCP throughput. Congestive to measure RTT along with measured TCP throughput. Congestive
skipping to change at page 10, line 48 skipping to change at page 10, line 48
This is not meant to provide an exhaustive list, but summarizes some This is not meant to provide an exhaustive list, but summarizes some
of the more common ways to determine round trip time (RTT) through of the more common ways to determine round trip time (RTT) through
the network. The desired resolution of the measurement (i.e. msec the network. The desired resolution of the measurement (i.e. msec
versus usec) may dictate whether the RTT measurement can be achieved versus usec) may dictate whether the RTT measurement can be achieved
with standard tools such as ICMP ping techniques or whether with standard tools such as ICMP ping techniques or whether
specialized test equipment would be required with high precision specialized test equipment would be required with high precision
timers. The objective in this section is to list several techniques timers. The objective in this section is to list several techniques
in order of decreasing accuracy. in order of decreasing accuracy.
- Use test equipment on each end of the network, "looping" the - Use test equipment on each end of the network, "looping" the
far-end tester so that a packet stream can be measured end-end. This far-end tester so that a packet stream can be measured end-end. This
test equipment RTT measurement may be compatible with delay test equipment RTT measurement may be compatible with delay
measurement protocols specified in RFC5357. measurement protocols specified in RFC5357.
- Conduct packet captures of TCP test applications using for example - Conduct packet captures of TCP test applications using for example
"iperf" or FTP, etc. By running multiple experiments, the packet "iperf" or FTP, etc. By running multiple experiments, the packet
captures can be studied to estimate RTT based upon the SYN -> SYN-ACK captures can be studied to estimate RTT based upon the SYN -> SYN-ACK
handshakes within the TCP connection set-up. handshakes within the TCP connection set-up.
- ICMP Pings may also be adequate to provide round trip time - ICMP Pings may also be adequate to provide round trip time
estimations. Some limitations of ICMP Ping are the msec resolution estimations. Some limitations of ICMP Ping are the msec resolution
and whether the network elements respond to pings (or block them). and whether the network elements respond to pings (or block them).
3.2.2 Techniques to Measure End-end Bandwidth 3.2.2 Techniques to Measure End-end Bandwidth
There are many well established techniques available to provide There are many well established techniques available to provide
estimated measures of bandwidth over a network. This measurement estimated measures of bandwidth over a network. This measurement
should be conducted in both directions of the network, especially for should be conducted in both directions of the network, especially for
access networks which are inherently asymmetrical. Some of the access networks which are inherently asymmetrical. Some of the
asymmetric implications to TCP performance are documented in RFC-3449 asymmetric implications to TCP performance are documented in RFC 3449
and the results of this work will be further studied to determine [RFC3449].
relevance to this draft.
The bandwidth measurement test must be run with stateless IP streams The bandwidth measurement test must be run with stateless IP streams
(not stateful TCP) in order to determine the available bandwidth in (not stateful TCP) in order to determine the available bandwidth in
each direction. And this test should obviously be performed at each direction. And this test should obviously be performed at
various intervals throughout a business day (or even across a week). various intervals throughout a business day (or even across a week).
Ideally, the bandwidth test should produce a log output of the Ideally, the bandwidth test should produce a log output of the
bandwidth achieved across the test interval AND the round trip delay. bandwidth achieved across the test interval AND the round trip delay.
And during the actual TCP level performance measurements (Sections And during the actual TCP level performance measurements (Sections
3.3 - 3.5), the test tool must be able to track round trip time 3.3 - 3.5), the test tool must be able to track round trip time
of the TCP connection(s) during the test. Measuring round trip time of the TCP connection(s) during the test. Measuring round trip time
variation (aka "jitter") provides insight into effects of congestive variation (aka "jitter") provides insight into effects of congestive
delay on the sustained throughput achieved for the TCP layer test. delay on the sustained throughput achieved for the TCP layer test.
3.3. TCP Throughput Tests 3.3. TCP Throughput Tests
This draft specifically defines TCP throughput techniques to verify This draft specifically defines TCP throughput techniques to verify
sustained TCP performance in a managed business network. Defined sustained TCP performance in a managed business network. Defined
in section 2.1, the equalibrium throughput reflects the maximum in section 2.1, the equilibrium throughput reflects the maximum
rate achieved by a TCP connection within the congestion avoidance rate achieved by a TCP connection within the congestion avoidance
phase on a end-end network path. This section and others will define phase on a end-end network path. This section and others will define
the method to conduct these sustained throughput tests and guidelines the method to conduct these sustained throughput tests and guidelines
of the predicted results. of the predicted results.
With baseline measurements of round trip time and bandwidth With baseline measurements of round trip time and bandwidth
from section 3.2, a series of single and multiple TCP connection from section 3.2, a series of single and multiple TCP connection
throughput tests can be conducted to baseline network performance throughput tests can be conducted to baseline network performance
against expectations. against expectations.
It is recommended to run the tests in each direction independently It is recommended to run the tests in each direction independently
first, then run both directions simultaneously. In each case, the TCP first, then run both directions simultaneously. In each case, the
Efficiency and TCP Transfer Time metrics must be measured in each TCP Efficiency and TCP Transfer Time metrics must be measured in each
direction. direction.
3.3.1 Calculate Optimum TCP Window Size 3.3.1 Calculate Optimum TCP Window Size
The optimum TCP window size can be calculated from the bandwidth delay The optimum TCP window size can be calculated from the bandwidth
product (BDP), which is: delay product (BDP), which is:
BDP (bits) = RTT (sec) x Bandwidth (bps) BDP (bits) = RTT (sec) x Bandwidth (bps)
By dividing the BDP by 8, the "ideal" TCP window size is calculated. By dividing the BDP by 8, the "ideal" TCP window size is calculated.
An example would be a T3 link with 25 msec RTT. The BDP would equal An example would be a T3 link with 25 msec RTT. The BDP would equal
~1,105,000 bits and the ideal TCP window would equal ~138,000 bytes. ~1,105,000 bits and the ideal TCP window would equal ~138,000 bytes.
The following table provides some representative network link speeds, The following table provides some representative network link speeds,
latency, BDP, and associated "optimum" TCP window size. Sustained latency, BDP, and associated "optimum" TCP window size. Sustained
TCP transfers should reach nearly 100% throughput, minus the overhead TCP transfers should reach nearly 100% throughput, minus the overhead
of Layers 1-3 and the divisor of the MSS into the window. of Layers 1-3 and the divisor of the MSS into the window.
For this single connection baseline test, the MSS size will effect For this single connection baseline test, the MSS size will effect
the achieved throughput (especially for smaller TCP window sizes). the achieved throughput (especially for smaller TCP window sizes).
Table 3.2 provides the achievable, equalibrium TCP throughput (at Table 3.2 provides the achievable, equilibrium TCP throughput (at
Layer 4) using 1460 byte MSS. Also in this table, the case of 58 byte Layer 4) using 1460 byte MSS. Also in this table, the 58 byte L1-L4
L1-L4 overhead including the Ethernet CRC32 is used for simplicity. overhead including the Ethernet CRC32 is used for simplicity.
Table 3.2: Link Speed, RTT and calculated BDP, TCP Throughput Table 3.2: Link Speed, RTT and calculated BDP, TCP Throughput
Link Ideal TCP Maximum Achievable Link Ideal TCP Maximum Achievable
Speed* RTT (ms) BDP (bits) Window (kbytes) TCP Throughput(Mbps) Speed* RTT (ms) BDP (bits) Window (kbytes) TCP Throughput(Mbps)
---------------------------------------------------------------------- ---------------------------------------------------------------------
T1 20 30,720 3.84 1.17 T1 20 30,720 3.84 1.17
T1 50 76,800 9.60 1.40 T1 50 76,800 9.60 1.40
T1 100 153,600 19.20 1.40 T1 100 153,600 19.20 1.40
T3 10 442,100 55.26 42.05 T3 10 442,100 55.26 42.05
T3 15 663,150 82.89 42.05 T3 15 663,150 82.89 42.05
T3 25 1,105,250 138.16 41.52 T3 25 1,105,250 138.16 41.52
T3(ATM) 10 407,040 50.88 36.50 T3(ATM) 10 407,040 50.88 36.50
T3(ATM) 15 610,560 76.32 36.23 T3(ATM) 15 610,560 76.32 36.23
T3(ATM) 25 1,017,600 127.20 36.27 T3(ATM) 25 1,017,600 127.20 36.27
100M 1 100,000 12.50 91.98 100M 1 100,000 12.50 91.98
100M 2 200,000 25.00 93.44 100M 2 200,000 25.00 93.44
100M 5 500,000 62.50 93.44 100M 5 500,000 62.50 93.44
1Gig 0.1 100,000 12.50 919.82 1Gig 0.1 100,000 12.50 919.82
1Gig 0.5 500,000 62.50 934.47 1Gig 0.5 500,000 62.50 934.47
1Gig 1 1,000,000 125.00 934.47 1Gig 1 1,000,000 125.00 934.47
10Gig 0.05 500,000 62.50 9,344.67 10Gig 0.05 500,000 62.50 9,344.67
10Gig 0.3 3,000,000 375.00 9,344.67 10Gig 0.3 3,000,000 375.00 9,344.67
* Note that link speed is the minimum link speed throughput a network; * Note that link speed is the minimum link speed throughput a
i.e. WAN with T1 link, etc. network; i.e. WAN with T1 link, etc.
Also, the following link speeds (available payload bandwidth) were Also, the following link speeds (available payload bandwidth) were
used for the WAN entries: used for the WAN entries:
- T1 = 1.536 Mbits/sec (B8ZS line encoding facility) - T1 = 1.536 Mbits/sec (B8ZS line encoding facility)
- T3 = 44.21 Mbits/sec (C-Bit Framing) - T3 = 44.21 Mbits/sec (C-Bit Framing)
- T3(ATM) = 36.86 Mbits/sec (C-Bit Framing & PLCP, 96000 Cells per - T3(ATM) = 36.86 Mbits/sec (C-Bit Framing & PLCP, 96000 Cells per
second) second)
The calculation method used in this document is a 3 step process : The calculation method used in this document is a 3 step process :
1 - We determine what should be the optimal TCP Window size value 1 - We determine what should be the optimal TCP Window size value
based on the optimal quantity of "in-flight" octets discovered by based on the optimal quantity of "in-flight" octets discovered by
the BDP calculation. We take into consideration that the TCP the BDP calculation. We take into consideration that the TCP
Window size has to be an exact multiple value of the MSS. Window size has to be an exact multiple value of the MSS.
2 - Then we calculate the achievable layer 2 throughput by multiplying 2 - Then we calculate the achievable layer 2 throughput by
the value determined in step 1 with the MSS & (MSS + L2 + L3 + L4 multiplying the value determined in step 1 with the
Overheads) divided by the RTT. MSS & (MSS + L2 + L3 + L4 Overheads) divided by the RTT.
3 - Finally, we multiply the calculated value of step 2 by the MSS 3 - Finally, we multiply the calculated value of step 2 by the MSS
versus (MSS + L2 + L3 + L4 Overheads) ratio. versus (MSS + L2 + L3 + L4 Overheads) ratio.
This gives us the achievable TCP Throughput value. Sometimes, the This gives us the achievable TCP Throughput value. Sometimes, the
maximum achievable throughput is limited by the maximum achievable maximum achievable throughput is limited by the maximum achievable
quantity of Ethernet Frames per second on the physical media. Then quantity of Ethernet Frames per second on the physical media. Then
this value is used in step 2 instead of the calculated one. this value is used in step 2 instead of the calculated one.
The following diagram compares achievable TCP throughputs on a T3 link The following diagram compares achievable TCP throughputs on a T3 link
with Windows 2000/XP TCP window sizes of 16KB versus 64KB. with Windows 2000/XP TCP window sizes of 16KB versus 64KB.
skipping to change at page 14, line 5 skipping to change at page 14, line 5
| | | | | |64K| | | | | | |64K|
15| 14.5M____| | | | | | 15| 14.5M____| | | | | |
| |16K| | | | | | | |16K| | | | | |
10| | | | 9.6M+---+ | | | 10| | | | 9.6M+---+ | | |
| | | | |16K| | 5.8M____+ | | | | | |16K| | 5.8M____+ |
5| | | | | | | |16K| | 5| | | | | | | |16K| |
|______+___+___+_______+___+___+_______+__ +___+_______ |______+___+___+_______+___+___+_______+__ +___+_______
10 15 25 10 15 25
RTT in milliseconds RTT in milliseconds
The following diagram shows the achievable TCP throughput on a 25ms T3 The following diagram shows the achievable TCP throughput on a 25ms
when the TCP Window size is increased and with the RFC1323 TCP Window T3 when the TCP Window size is increased and with the RFC1323 TCP
scaling option. Window scaling option.
45| 45|
| +-----+42.47M | +-----+42.47M
40| | | 40| | |
TCP | | | TCP | | |
Throughput 35| | | Throughput 35| | |
in Mbps | | | in Mbps | | |
30| | | 30| | |
| | | | | |
25| | | 25| | |
| ______ 21.23M | | | ______ 21.23M | |
20| | | | | 20| | | | |
| | | | | | | | | |
15| | | | | 15| | | | |
| | | | | | | | | |
10| +----+10.62M | | | | 10| +----+10.62M | | | |
| _______5.31M | | | | | | | _______5.31M | | | | | |
5| | | | | | | | | 5| | | | | | | | |
|__+_____+______+____+___________+____+________+_____+___ |__+_____+______+____+__________+____+________+_____+___
16 32 64 128 16 32 64 128
TCP Window size in KBytes TCP Window size in KBytes
3.3.2 Conducting the TCP Throughput Tests 3.3.2 Conducting the TCP Throughput Tests
There are several TCP tools that are commonly used in the network There are several TCP tools that are commonly used in the network
world and one of the most common is the "iperf" tool. With this tool, world and one of the most common is the "iperf" tool. With this tool,
hosts are installed at each end of the network segment; one as client hosts are installed at each end of the network segment; one as client
and the other as server. The TCP Window size of both the client and and the other as server. The TCP Window size of both the client and
the server can be maunally set and the achieved throughput is measured, the server can be manually set and the achieved throughput is
either uni-directionally or bi-directionally. For higher BDP measured, either uni-directionally or bi-directionally. For higher
situations in lossy networks (long fat networks or satellite links, BDP situations in lossy networks (long fat networks or satellite
etc.), TCP options such as Selective Acknowledgment should be links, etc.), TCP options such as Selective Acknowledgment should be
considered and also become part of the window size / throughput considered and also become part of the window size / throughput
characterization. characterization.
Host hardware performance must be well understood before conducting Host hardware performance must be well understood before conducting
the TCP throughput tests and other tests in the following sections. the TCP throughput tests and other tests in the following sections.
Dedicated test equipment will generally be required, especially for Dedicated test equipment will generally be required, especially for
line rates of GigE and 10 GigE. line rates of GigE and 10 GigE.
The TCP throughput test should be run over a a long enough duration The TCP throughput test should be run over a a long enough duration
to properly exercise network buffers and also characterize performance to properly exercise network buffers and also characterize
during different time periods of the day. The results must be logged performance during different time periods of the day. The results
at the desired interval and the test must record RTT and TCP must be logged at the desired interval and the test must record RTT
retransmissions at each interval. and TCP retransmissions at each interval.
This correlation of retransmissions and RTT over the course of the This correlation of retransmissions and RTT over the course of the
test will clearly identify which portions of the transfer reached test will clearly identify which portions of the transfer reached
TCP Equilbrium state and to what effect increased RTT (congestive TCP Equilbrium state and to what effect increased RTT (congestive
effects) may have been the cause of reduced equilibrium performance. effects) may have been the cause of reduced equilibrium performance.
Additionally, the TCP Efficiency and TCP Transfer time metrics should Additionally, the TCP Efficiency and TCP Transfer time metrics should
be logged in order to further characterize the window size tests. be logged in order to further characterize the window size tests.
3.3.3 Single vs. Multiple TCP Connection Testing 3.3.3 Single vs. Multiple TCP Connection Testing
The decision whether to conduct single or multiple TCP connection The decision whether to conduct single or multiple TCP connection
tests depends upon the size of the BDP in relation to the window sizes tests depends upon the size of the BDP in relation to the window
configured in the end-user environment. For example, if the BDP for a sizes configured in the end-user environment. For example, if the
long-fat pipe turns out to be 2MB, then it is probably more realistic BDP for a long-fat pipe turns out to be 2MB, then it is probably more
to test this pipe with multiple connections. Assuming typical host realistic to test this pipe with multiple connections. Assuming
computer window settings of 64 KB, using 32 connections would typical host computer window settings of 64 KB, using 32 connections
realistically test this pipe. would realistically test this pipe.
The following table is provided to illustrate the relationship of the The following table is provided to illustrate the relationship of the
BDP, window size, and the number of connections required to utilize the BDP, window size, and the number of connections required to utilize
the available capacity. For this example, the network bandwidth is the available capacity. For this example, the network bandwidth is
500 Mbps, RTT is equal to 5 ms, and the BDP equates to 312 KBytes. 500 Mbps, RTT is equal to 5 ms, and the BDP equates to 312 KBytes.
#Connections #Connections
Window to Fill Link Window to Fill Link
------------------------ ------------------------
16KB 20 16KB 20
32KB 10 32KB 10
64KB 5 64KB 5
128KB 3 128KB 3
The TCP Transfer Time metric is useful for conducting multiple The TCP Transfer Time metric is useful for conducting multiple
connection tests. Each connection should be configured to transfer connection tests. Each connection should be configured to transfer
a certain payload (i.e. 100 MB), and the TCP Transfer time provides a certain payload (i.e. 100 MB), and the TCP Transfer time provides
a simple metric to verify the actual versus expected results. a simple metric to verify the actual versus expected results.
Note that the TCP transfer time is the time for all connections to Note that the TCP transfer time is the time for all connections to
complete the transfer of the configured payload size. From the complete the transfer of the configured payload size. From the
example table listed above, the 64KB window is considered. Each of example table listed above, the 64KB window is considered. Each of
the 5 connections would be configured to transfer 100MB, and each the 5 connections would be configured to transfer 100MB, and each
TCP should obtain a maximum of 100 Mb/sec per connection. So for this TCP should obtain a maximum of 100 Mb/sec per connection. So for
example, the 100MB payload should be transferred across the connections this example, the 100MB payload should be transferred across the
in approximately 8 seconds (which would be the ideal TCP transfer time connections in approximately 8 seconds (which would be the ideal TCP
for these conditions). transfer time for these conditions).
Additionally, the TCP Efficiency metric should be computed for each Additionally, the TCP Efficiency metric should be computed for each
connection tested (defined in section 2.2). connection tested (defined in section 2.2).
3.3.4 Interpretation of the TCP Throughput Results 3.3.4 Interpretation of the TCP Throughput Results
At the end of this step, the user will document the theoretical BDP At the end of this step, the user will document the theoretical BDP
and a set of Window size experiments with measured TCP throughput for and a set of Window size experiments with measured TCP throughput for
each TCP window size setting. For cases where the sustained TCP each TCP window size setting. For cases where the sustained TCP
throughput does not equal the predicted value, some possible causes throughput does not equal the predicted value, some possible causes
are listed: are listed:
- Network congestion causing packet loss; the TCP Efficiency metric - Network congestion causing packet loss; the TCP Efficiency metric
is a useful gauge to compare network performance is a useful gauge to compare network performance
- Network congestion not causing packet loss but increasing RTT - Network congestion not causing packet loss but increasing RTT
- Intermediate network devices which actively regenerate the TCP - Intermediate network devices which actively regenerate the TCP
connection and can alter window size, MSS, etc. connection and can alter window size, MSS, etc.
- Over utilization of available link or rate limiting (policing). More - Over utilization of available link or rate limiting (policing).
discussion of traffic management tests follows in section 3.4 More discussion of traffic management tests follows in section 3.4
3.4. Traffic Management Tests 3.4. Traffic Management Tests
In most cases, the network connection between two geographic locations In most cases, the network connection between two geographic
(branch offices, etc.) is lower than the network connection of the locations (branch offices, etc.) is lower than the network connection
host computers. An example would be LAN connectivity of GigE and of the host computers. An example would be LAN connectivity of GigE
WAN connectivity of 100 Mbps. The WAN connectivity may be physically and WAN connectivity of 100 Mbps. The WAN connectivity may be
100 Mbps or logically 100 Mbps (over a GigE WAN connection). In the physically 100 Mbps or logically 100 Mbps (over a GigE WAN
later case, rate limiting is used to provide the WAN bandwidth per the connection). In the later case, rate limiting is used to provide the
SLA. WAN bandwidth per the SLA.
Traffic management techniques are employed to provide various forms of Traffic management techniques are employed to provide various forms
QoS, the more common include: of QoS, the more common include:
- Traffic Shaping - Traffic Shaping
- Priority Queuing - Priority Queueing
- Random Early Discard (RED, etc.) - Random Early Discard (RED, etc.)
Configuring the end-end network with these various traffic management Configuring the end-end network with these various traffic management
mechanisms is a complex under-taking. For traffic shaping and RED mechanisms is a complex under-taking. For traffic shaping and RED
techniques, the end goal is to provide better performance for bursty techniques, the end goal is to provide better performance for bursty
traffic such as TCP (RED is specifically intended for TCP). traffic such as TCP (RED is specifically intended for TCP).
This section of the methodology provides guidelines to test traffic This section of the methodology provides guidelines to test traffic
shaping and RED implementations. As in section 3.3, host hardware shaping and RED implementations. As in section 3.3, host hardware
performance must be well understood before conducting the traffic performance must be well understood before conducting the traffic
skipping to change at page 17, line 14 skipping to change at page 17, line 14
Simply stated, traffic policing marks and/or drops packets which Simply stated, traffic policing marks and/or drops packets which
exceed the SLA bandwidth (in most cases, excess traffic is dropped). exceed the SLA bandwidth (in most cases, excess traffic is dropped).
Traffic shaping employs the use of queues to smooth the bursty Traffic shaping employs the use of queues to smooth the bursty
traffic and then send out within the SLA bandwidth limit (without traffic and then send out within the SLA bandwidth limit (without
dropping packets unless the traffic shaping queue is exceeded). dropping packets unless the traffic shaping queue is exceeded).
Traffic shaping is generally configured for TCP data services and Traffic shaping is generally configured for TCP data services and
can provide improved TCP performance since the retransmissions are can provide improved TCP performance since the retransmissions are
reduced, which in turn optimizes TCP throughput for the given reduced, which in turn optimizes TCP throughput for the given
available bandwidth. Through this section, the available rate-limited available bandwidth. Through this section, the available
bandwidth shall be referred to as the "bottleneck bandwidth". rate-limited bandwidth shall be referred to as the
"bottleneck bandwidth".
The ability to detect proper traffic shaping is more easily diagnosed The ability to detect proper traffic shaping is more easily diagnosed
when conducting a multiple TCP connection test. Proper shaping will when conducting a multiple TCP connection test. Proper shaping will
provide a fair distribution of the available bottleneck bandwidth, provide a fair distribution of the available bottleneck bandwidth,
while traffic policing will not. while traffic policing will not.
The traffic shaping tests build upon the concepts of multiple The traffic shaping tests build upon the concepts of multiple
connection testing as defined in section 3.3.3. Calculating the BDP connection testing as defined in section 3.3.3. Calculating the BDP
for the bottleneck bandwidth is first required and then selecting for the bottleneck bandwidth is first required and then selecting
the number of connections / window size per connection. the number of connections / window size per connection.
Similar to the example in section 3.3, a typical test scenario might Similar to the example in section 3.3, a typical test scenario might
be: GigE LAN with a 100Mbps bottleneck bandwidth (rate limited logical be: GigE LAN with a 100Mbps bottleneck bandwidth (rate limited
interface), and 5 msec RTT. This would require five (5) TCP logical interface), and 5 msec RTT. This would require five (5) TCP
connections of 64 KB window size evenly fill the bottleneck bandwidth connections of 64 KB window size evenly fill the bottleneck bandwidth
(about 100 Mbps per connection). (about 100 Mbps per connection).
The traffic shaping should be run over a long enough duration to The traffic shaping should be run over a long enough duration to
properly exercise network buffers and also characterize performance properly exercise network buffers and also characterize performance
during different time periods of the day. The throughput of each during different time periods of the day. The throughput of each
connection must be logged during the entire test, along with the TCP connection must be logged during the entire test, along with the TCP
Efficiency and TCP Transfer time metric. Additionally, it is Efficiency and TCP Transfer time metric. Additionally, it is
recommended to log RTT and retransmissions per connection over the test recommended to log RTT and retransmissions per connection over the
interval. test interval.
3.4.1.1 Interpretation of Traffic Shaping Test Restults 3.4.1.1 Interpretation of Traffic Shaping Test Restults
By plotting the throughput achieved by each TCP connection, the fair By plotting the throughput achieved by each TCP connection, the fair
sharing of the bandwidth is generally very obvious when traffic shaping sharing of the bandwidth is generally very obvious when traffic
is properly configured for the bottleneck interface. For the previous shaping is properly configured for the bottleneck interface. For the
example of 5 connections sharing 500 Mbps, each connection would previous example of 5 connections sharing 500 Mbps, each connection
consume ~100 Mbps with a smooth variation. If traffic policing was would consume ~100 Mbps with a smooth variation. If traffic policing
present on the bottleneck interface, the bandwidth sharing would not was present on the bottleneck interface, the bandwidth sharing would
be fair and the resulting throughput plot would reveal "spikey" not be fair and the resulting throughput plot would reveal "spikey"
connection throughput consumption of the competing TCP connections throughput consumption of the competing TCP connections (due to the
(due to the retransmissions). retransmissions).
3.4.2 RED Tests 3.4.2 RED Tests
Random Early Discard techniques are specifically targeted to provide Random Early Discard techniques are specifically targeted to provide
congestion avoidance for TCP traffic. Before the network element queue congestion avoidance for TCP traffic. Before the network element
"fills" and enters the tail drop state, RED drops packets at queue "fills" and enters the tail drop state, RED drops packets at
configurable queue depth thresholds. This action causes TCP configurable queue depth thresholds. This action causes TCP
connections to back-off which helps to prevent tail drop, which in connections to back-off which helps to prevent tail drop, which in
turn helps to prevent global TCP synchronization. turn helps to prevent global TCP synchronization.
Again, rate limited interfaces can benefit greatly from RED based Again, rate limited interfaces can benefit greatly from RED based
techniques. Without RED, TCP is generally not able to achieve the full techniques. Without RED, TCP is generally not able to achieve the
bandwidth of the bottleneck interface. With RED enabled, TCP full bandwidth of the bottleneck interface. With RED enabled, TCP
congestion avoidance throttles the connections on the higher speed congestion avoidance throttles the connections on the higher speed
interface (i.e. LAN) and can reach equalibrium with the bottleneck interface (i.e. LAN) and can reach equalibrium with the bottleneck
bandwidth (achieving closer to full throughput). bandwidth (achieving closer to full throughput).
The ability to detect proper RED configuration is more easily diagnosed The ability to detect proper RED configuration is more easily
when conducting a multiple TCP connection test. Multiple TCP diagnosed when conducting a multiple TCP connection test. Multiple
connections provide the multiple bursty sources that emulate the TCP connections provide the multiple bursty sources that emulate the
real-world conditions for which RED was intended. real-world conditions for which RED was intended.
The RED tests also build upon the concepts of multiple connection The RED tests also build upon the concepts of multiple connection
testing as defined in secion 3.3.3. Calculating the BDP for the testing as defined in secion 3.3.3. Calculating the BDP for the
bottleneck bandwidth is first required and then selecting the number of bottleneck bandwidth is first required and then selecting the number
connections / window size per connection. of connections / window size per connection.
For RED testing, the desired effect is to cause the TCP connections to For RED testing, the desired effect is to cause the TCP connections
burst beyond the bottleneck bandwidth so that queue drops will occur. to burst beyond the bottleneck bandwidth so that queue drops will
Using the same example from section 3.4.1 (traffic shaping), the occur. Using the same example from section 3.4.1 (traffic shaping),
500 Mbps bottleneck bandwidth requires 5 TCP connections (with window the 500 Mbps bottleneck bandwidth requires 5 TCP connections (with
size of 64Kb) to fill the capacity. Some experimentation is required, window size of 64Kb) to fill the capacity. Some experimentation is
but it is recommended to start with double the number of connections required,but it is recommended to start with double the number of
to stress the network element buffers / queues. In this example, 10 connections to stress the network element buffers / queues. In this
connections would produce TCP bursts of 64KB for each connection. example, 10 connections would produce TCP bursts of 64KB for each
If the timing of the TCP tester permits, these TCP bursts could stress connection. If the timing of the TCP tester permits, these TCP
queue sizes in the 512KB range. Again experimentation will be required bursts could stress queue sizes in the 512KB range. Again
and the proper number of TCP connections / window size will be dictated experimentation will be required and the proper number of TCP
by the size the network element queue. connections / window size will be dictated by the size the network
element queue.
3.4.2.1 Interpretation of RED Results 3.4.2.1 Interpretation of RED Results
The default queuing technique for most network devices is FIFO based. The default queuing technique for most network devices is FIFO based.
Without RED, the FIFO based queue will cause excessive loss to all of Without RED, the FIFO based queue will cause excessive loss to all of
the TCP connections and in the worst case global TCP synchronization. the TCP connections and in the worst case global TCP synchronization.
By plotting the aggregate throughput achieved on the bottleneck By plotting the aggregate throughput achieved on the bottleneck
interface, proper RED operation can be determined if the bottleneck interface, proper RED operation can be determined if the bottleneck
bandwidth is fully utilized. For the previous example of 10 bandwidth is fully utilized. For the previous example of 10
connections (window = 64 KB) sharing 500 Mbps, each connection should connections (window = 64 KB) sharing 500 Mbps, each connection should
consume ~50 Mbps. If RED was not properly enabled on the interface, consume ~50 Mbps. If RED was not properly enabled on the interface,
then the TCP connections will retransmit at a higher rate and the net then the TCP connections will retransmit at a higher rate and the net
effect is that the bottleneck bandwidth is not fully utilized. effect is that the bottleneck bandwidth is not fully utilized.
Another means to study non-RED versus RED implementation is to use Another means to study non-RED versus RED implementation is to use
the TCP Transfer Time metric for all of the connections. In this the TCP Transfer Time metric for all of the connections. In this
example, a 100 MB payload transfer should take ideally 16 seconds example, a 100 MB payload transfer should take ideally 16 seconds
across all 10 connections (with RED enabled). With RED not enabled, across all 10 connections (with RED enabled). With RED not enabled,
the throughput across the bottleneck bandwidth would be greatly reduced the throughput across the bottleneck bandwidth would be greatly
(generally 20-40%) and the TCP Transfer time would be proportionally reduced (generally 20-40%) and the TCP Transfer time would be
longer then the ideal transfer time. proportionally longer then the ideal transfer time.
Additionally, the TCP Transfer Efficiency metric is useful, since Additionally, the TCP Transfer Efficiency metric is useful, since
non-RED implementations will exhibit a lower TCP Tranfer Efficiency non-RED implementations will exhibit a lower TCP Tranfer Efficiency
than RED implementations. than RED implementations.
4. Acknowledgements 4. Security Considerations
The security considerations that apply to any active measurement of
live networks are relevant here as well. See [RFC4656] and
[RFC5357].
5. IANA Considerations
This memo does not require and IANA registration for ports dedicated
to the TCP testing described in this memo.
6. Acknowledgements
The author would like to thank Gilles Forget, Loki Jorgenson, The author would like to thank Gilles Forget, Loki Jorgenson,
and Reinhard Schrage for technical review and original contributions and Reinhard Schrage for technical review and original contributions
to this draft-03. to this draft-06.
Also thanks to Matt Mathis and Matt Zekauskas for many good comments Also thanks to Matt Mathis, Matt Zekauskas, Al Morton, and Yaakov
through email exchange and for pointing us to great sources of Stein for many good comments and for pointing us to great sources of
information pertaining to past works in the TCP capacity area. information pertaining to past works in the TCP capacity area.
5. References 7. References
[RFC2581] Allman, M., Paxson, V., Stevens W., "TCP Congestion 7.1 Normative References
Control", RFC 2581, June 1999.
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC4656] Shalunov, S., Teitelbaum, B., Karp, A., Boote, J., and M.
Zekauskas, "A One-way Active Measurement Protocol
(OWAMP)", RFC 4656, September 2006.
[RFC5681] Allman, M., Paxson, V., Stevens W., "TCP Congestion
Control", RFC 5681, September 2009.
[RFC3148] Mathis M., Allman, M., "A Framework for Defining
Empirical Bulk Transfer Capacity Metrics", RFC 3148, July
2001.
[RFC2544] Bradner, S., McQuaid, J., "Benchmarking Methodology for [RFC2544] Bradner, S., McQuaid, J., "Benchmarking Methodology for
Network Interconnect Devices", RFC 2544, June 1999 Network Interconnect Devices", RFC 2544, June 1999
[RFC3449] Balakrishnan, H., Padmanabhan, V. N., Fairhurst, G., [RFC3449] Balakrishnan, H., Padmanabhan, V. N., Fairhurst, G.,
Sooriyabandara, M., "TCP Performance Implications of Sooriyabandara, M., "TCP Performance Implications of
Network Path Asymmetry", RFC 3449, December 2002 Network Path Asymmetry", RFC 3449, December 2002
[RFC5357] Hedayat, K., Krzanowski, R., Morton, A., Yum, K., Babiarz, [RFC5357] Hedayat, K., Krzanowski, R., Morton, A., Yum, K., Babiarz,
J., "A Two-Way Active Measurement Protocol (TWAMP)", J., "A Two-Way Active Measurement Protocol (TWAMP)",
RFC 5357, October 2008 RFC 5357, October 2008
skipping to change at page 20, line 4 skipping to change at page 20, line 21
[RFC3449] Balakrishnan, H., Padmanabhan, V. N., Fairhurst, G., [RFC3449] Balakrishnan, H., Padmanabhan, V. N., Fairhurst, G.,
Sooriyabandara, M., "TCP Performance Implications of Sooriyabandara, M., "TCP Performance Implications of
Network Path Asymmetry", RFC 3449, December 2002 Network Path Asymmetry", RFC 3449, December 2002
[RFC5357] Hedayat, K., Krzanowski, R., Morton, A., Yum, K., Babiarz, [RFC5357] Hedayat, K., Krzanowski, R., Morton, A., Yum, K., Babiarz,
J., "A Two-Way Active Measurement Protocol (TWAMP)", J., "A Two-Way Active Measurement Protocol (TWAMP)",
RFC 5357, October 2008 RFC 5357, October 2008
[RFC4821] Mathis, M., Heffner, J., "Packetization Layer Path MTU [RFC4821] Mathis, M., Heffner, J., "Packetization Layer Path MTU
Discovery", RFC 4821, June 2007 Discovery", RFC 4821, June 2007
draft-ietf-ippm-btc-cap-00.txt Allman, M., "A Bulk draft-ietf-ippm-btc-cap-00.txt Allman, M., "A Bulk
Transfer Capacity Methodology for Cooperating Hosts", Transfer Capacity Methodology for Cooperating Hosts",
August 2001 August 2001
[MSMO] The Macroscopic Behavior of the TCP Congestion Avoidance 7.2. Informative References
Algorithm Mathis, M.,Semke, J, Mahdavi, J, Ott, T
July 1997 SIGCOMM Computer Communication Review,
Volume 27 Issue 3
[Stevens Vol1] TCP/IP Illustrated, Vol1, The Protocols
Addison-Wesley
Authors' Addresses Authors' Addresses
Barry Constantine Barry Constantine
JDSU, Test and Measurement Division JDSU, Test and Measurement Division
One Milesone Center Court One Milesone Center Court
Germantown, MD 20876-7100 Germantown, MD 20876-7100
USA USA
Phone: +1 240 404 2227 Phone: +1 240 404 2227
 End of changes. 75 change blocks. 
214 lines changed or deleted 238 lines changed or added

This html diff was produced by rfcdiff 1.38. The latest version is available from http://tools.ietf.org/tools/rfcdiff/