rfc9406.original   rfc9406.txt 
Network Working Group P. Balasubramanian Internet Engineering Task Force (IETF) P. Balasubramanian
Internet-Draft Confluent Request for Comments: 9406 Confluent
Intended status: Standards Track Y. Huang Category: Standards Track Y. Huang
Expires: 31 August 2023 M. Olson ISSN: 2070-1721 M. Olson
Microsoft Microsoft
27 February 2023 May 2023
HyStart++: Modified Slow Start for TCP HyStart++: Modified Slow Start for TCP
draft-ietf-tcpm-hystartplusplus-14
Abstract Abstract
This document describes HyStart++, a simple modification to the slow This document describes HyStart++, a simple modification to the slow
start phase of congestion control algorithms. Slow start can start phase of congestion control algorithms. Slow start can
overshoot the ideal send rate in many cases, causing high packet loss overshoot the ideal send rate in many cases, causing high packet loss
and poor performance. HyStart++ uses increase in round-trip delay as and poor performance. HyStart++ uses increase in round-trip delay as
a heuristic to find an exit point before possible overshoot. It also a heuristic to find an exit point before possible overshoot. It also
adds a mitigation to prevent jitter from causing premature slow start adds a mitigation to prevent jitter from causing premature slow start
exit. exit.
Status of This Memo Status of This Memo
This Internet-Draft is submitted in full conformance with the This is an Internet Standards Track document.
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months This document is a product of the Internet Engineering Task Force
and may be updated, replaced, or obsoleted by other documents at any (IETF). It represents the consensus of the IETF community. It has
time. It is inappropriate to use Internet-Drafts as reference received public review and has been approved for publication by the
material or to cite them other than as "work in progress." Internet Engineering Steering Group (IESG). Further information on
Internet Standards is available in Section 2 of RFC 7841.
This Internet-Draft will expire on 31 August 2023. Information about the current status of this document, any errata,
and how to provide feedback on it may be obtained at
https://www.rfc-editor.org/info/rfc9406.
Copyright Notice Copyright Notice
Copyright (c) 2023 IETF Trust and the persons identified as the Copyright (c) 2023 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 (https://trustee.ietf.org/ Provisions Relating to IETF Documents
license-info) in effect on the date of publication of this document. (https://trustee.ietf.org/license-info) in effect on the date of
Please review these documents carefully, as they describe your rights publication of this document. Please review these documents
and restrictions with respect to this document. Code Components carefully, as they describe your rights and restrictions with respect
extracted from this document must include Revised BSD License text as to this document. Code Components extracted from this document must
described in Section 4.e of the Trust Legal Provisions and are include Revised BSD License text as described in Section 4.e of the
provided without warranty as described in the Revised BSD License. Trust Legal Provisions and are provided without warranty as described
in the Revised BSD License.
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 1. Introduction
2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. Terminology
3. Definitions . . . . . . . . . . . . . . . . . . . . . . . . . 3 3. Definitions
4. HyStart++ Algorithm . . . . . . . . . . . . . . . . . . . . . 3 4. HyStart++ Algorithm
4.1. Summary . . . . . . . . . . . . . . . . . . . . . . . . . 3 4.1. Summary
4.2. Algorithm Details . . . . . . . . . . . . . . . . . . . . 4 4.2. Algorithm Details
4.3. Tuning constants and other considerations . . . . . . . . 6 4.3. Tuning Constants and Other Considerations
5. Deployments and Performance Evaluations . . . . . . . . . . . 7 5. Deployments and Performance Evaluations
6. Security Considerations . . . . . . . . . . . . . . . . . . . 8 6. Security Considerations
7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 8 7. IANA Considerations
8. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 8 8. References
9. References . . . . . . . . . . . . . . . . . . . . . . . . . 8 8.1. Normative References
9.1. Normative References . . . . . . . . . . . . . . . . . . 8 8.2. Informative References
9.2. Informative References . . . . . . . . . . . . . . . . . 8 Acknowledgments
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 9 Authors' Addresses
1. Introduction 1. Introduction
[RFC5681] describes the slow start congestion control algorithm for [RFC5681] describes the slow start congestion control algorithm for
TCP. The slow start algorithm is used when the congestion window TCP. The slow start algorithm is used when the congestion window
(cwnd) is less than the slow start threshold (ssthresh). During slow (cwnd) is less than the slow start threshold (ssthresh). During slow
start, in absence of packet loss signals, TCP increases cwnd start, in the absence of packet loss signals, TCP increases the cwnd
exponentially to probe the network capacity. This fast growth can exponentially to probe the network capacity. This fast growth can
overshoot the ideal sending rate and cause significant packet loss overshoot the ideal sending rate and cause significant packet loss
which cannot always be recovered efficiently. that cannot always be recovered efficiently.
HyStart++ uses increase in round-trip delay as a signal to exit slow HyStart++ builds upon Hybrid Start (HyStart), originally described in
start before potential packet loss occurs as a result of overshoot. [HyStart]. HyStart++ uses increase in round-trip delay as a signal
This is one of two algorithms specified in [HyStart]. After the slow to exit slow start before potential packet loss occurs as a result of
start exit, a new Conservative Slow Start (CSS) phase is used to overshoot. This is one of two algorithms specified in [HyStart] for
determine whether the slow start exit was premature and to resume finding a safe exit point for slow start. After the slow start exit,
slow start. This mitigation improves performance in presence of a new Conservative Slow Start (CSS) phase is used to determine
jitter. HyStart++ reduces packet loss and retransmissions, and whether the slow start exit was premature and to resume slow start.
improves goodput in lab measurements and real world deployments. This mitigation improves performance in the presence of jitter.
HyStart++ reduces packet loss and retransmissions, and improves
goodput in lab measurements and real-world deployments.
While this document describes Hystart++ for TCP, it can also be used While this document describes HyStart++ for TCP, it can also be used
for other transport protocols which use slow start such as QUIC for other transport protocols that use slow start, such as QUIC
[RFC9002] or SCTP [RFC9260]. [RFC9002] or the Stream Control Transmission Protocol (SCTP)
[RFC9260].
2. Terminology 2. Terminology
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in BCP "OPTIONAL" in this document are to be interpreted as described in
14 [RFC2119] [RFC8174] when, and only when, they appear in all BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all
capitals, as shown here. capitals, as shown here.
3. Definitions 3. Definitions
We repeat here some definition from [RFC5681] to aid the reader. To aid the reader, we repeat some definitions from [RFC5681]:
SENDER MAXIMUM SEGMENT SIZE (SMSS): The SMSS is the size of the SENDER MAXIMUM SEGMENT SIZE (SMSS): The size of the largest segment
largest segment that the sender can transmit. This value can be that the sender can transmit. This value can be based on the
based on the maximum transmission unit of the network, the path MTU maximum transmission unit of the network, the Path MTU Discovery
discovery [RFC1191], [RFC4821] algorithm, RMSS (see next item), or algorithm [RFC1191] [RFC4821], RMSS (see next item), or other
other factors. The size does not include the TCP/IP headers and factors. The size does not include the TCP/IP headers and
options. options.
RECEIVER MAXIMUM SEGMENT SIZE (RMSS): The RMSS is the size of the RECEIVER MAXIMUM SEGMENT SIZE (RMSS): The size of the largest
largest segment the receiver is willing to accept. This is the value segment that the receiver is willing to accept. This is the value
specified in the MSS option sent by the receiver during connection specified in the MSS option sent by the receiver during connection
startup. Or, if the MSS option is not used, it is 536 bytes startup. Or, if the MSS option is not used, it is 536 bytes
[RFC1122]. The size does not include the TCP/IP headers and options. [RFC1122]. The size does not include the TCP/IP headers and
options.
RECEIVER WINDOW (rwnd): The most recently advertised receiver window. RECEIVER WINDOW (rwnd): The most recently advertised receiver
window.
CONGESTION WINDOW (cwnd): A TCP state variable that limits the amount CONGESTION WINDOW (cwnd): A TCP state variable that limits the
of data a TCP can send. At any given time, a TCP MUST NOT send data amount of data a TCP can send. At any given time, a TCP MUST NOT
with a sequence number higher than the sum of the highest send data with a sequence number higher than the sum of the
acknowledged sequence number and the minimum of cwnd and rwnd. highest acknowledged sequence number and the minimum of the cwnd
and rwnd.
4. HyStart++ Algorithm 4. HyStart++ Algorithm
4.1. Summary 4.1. Summary
[HyStart] specifies two algorithms (a "Delay Increase" algorithm and [HyStart] specifies two algorithms (a "Delay Increase" algorithm and
an "Inter-Packet Arrival" algorithm) to be run in parallel to detect an "Inter-Packet Arrival" algorithm) to be run in parallel to detect
that the sending rate has reached capacity. In practice, the Inter- that the sending rate has reached capacity. In practice, the Inter-
Packet Arrival algorithm does not perform well and is not able to Packet Arrival algorithm does not perform well and is not able to
detect congestion early, primarily due to ACK compression. The idea detect congestion early, primarily due to ACK compression. The idea
of the Delay Increase algorithm is to look for spikes in RTT (round- of the Delay Increase algorithm is to look for spikes in RTT (round-
trip time), which suggest that the bottleneck buffer is filling up. trip time), which suggest that the bottleneck buffer is filling up.
In HyStart++, a TCP sender uses traditional slow start and then uses In HyStart++, a TCP sender uses standard slow start and then uses the
the "Delay Increase" algorithm to trigger an exit from slow start. Delay Increase algorithm to trigger an exit from slow start. But
But instead of going straight from slow start to congestion instead of going straight from slow start to congestion avoidance,
avoidance, the sender spends a number of RTTs in a Conservative Slow the sender spends a number of RTTs in a Conservative Slow Start (CSS)
Start (CSS) phase to determine whether the exit from slow start was phase to determine whether the exit from slow start was premature.
premature. During CSS, the congestion window is grown exponentially During CSS, the congestion window is grown exponentially in a fashion
like in regular slow start, but with a smaller exponential base, similar to regular slow start, but with a smaller exponential base,
resulting in less aggressive growth. If the RTT reduces during CSS, resulting in less aggressive growth. If the RTT reduces during CSS,
it's concluded that the RTT spike was not related to congestion it's concluded that the RTT spike was not related to congestion
caused by the connection sending at a rate greater than the ideal caused by the connection sending at a rate greater than the ideal
send rate, and the connection resumes slow start. If the RTT send rate, and the connection resumes slow start. If the RTT
inflation persists throughout CSS, the connection enters congestion inflation persists throughout CSS, the connection enters congestion
avoidance. avoidance.
4.2. Algorithm Details 4.2. Algorithm Details
The following pseudocode uses a limit, L, to control the The following pseudocode uses a limit, L, to control the
aggressiveness of the cwnd increase during both standard slow start aggressiveness of the cwnd increase during both standard slow start
and CSS. While an arriving ACK may newly acknowledge an arbitrary and CSS. While an arriving ACK may newly acknowledge an arbitrary
number of bytes, the Hystart++ algorithm limits the number of those number of bytes, the HyStart++ algorithm limits the number of those
bytes applied to increase the cwnd to L*SMSS bytes. bytes applied to increase the cwnd to L*SMSS bytes.
lastRoundMinRTT and currentRoundMinRTT are initialized to infinity at lastRoundMinRTT and currentRoundMinRTT are initialized to infinity at
the initialization time. currRTT is the RTT sampled from the latest the initialization time. currRTT is the RTT sampled from the latest
incoming ACK and initialized to infinity. incoming ACK and initialized to infinity.
lastRoundMinRTT = infinity lastRoundMinRTT = infinity
currentRoundMinRTT = infinity currentRoundMinRTT = infinity
currRTT = infinity currRTT = infinity
Hystart++ measures rounds using sequence numbers, as follows: Define HyStart++ measures rounds using sequence numbers, as follows:
windowEnd as a sequence number initialized to SND.NXT. When
windowEnd is ACKed, the current round ends and windowEnd is set to
SND.NXT.
At the start of each round during standard slow start ([RFC5681]) and * Define windowEnd as a sequence number initialized to SND.NXT.
CSS, initialize the variables used to compute last round and current
round's minimum RTT: * When windowEnd is ACKed, the current round ends and windowEnd is
set to SND.NXT.
At the start of each round during standard slow start [RFC5681] and
CSS, initialize the variables used to compute the last round's and
current round's minimum RTT:
lastRoundMinRTT = currentRoundMinRTT lastRoundMinRTT = currentRoundMinRTT
currentRoundMinRTT = infinity currentRoundMinRTT = infinity
rttSampleCount = 0 rttSampleCount = 0
For each arriving ACK in slow start, where N is the number of For each arriving ACK in slow start, where N is the number of
previously unacknowledged bytes acknowledged in the arriving ACK: previously unacknowledged bytes acknowledged in the arriving ACK:
Update the cwnd: Update the cwnd:
cwnd = cwnd + min(N, L * SMSS) cwnd = cwnd + min(N, L * SMSS)
Keep track of minimum observed RTT: Keep track of the minimum observed RTT:
currentRoundMinRTT = min(currentRoundMinRTT, currRTT) currentRoundMinRTT = min(currentRoundMinRTT, currRTT)
rttSampleCount += 1 rttSampleCount += 1
For rounds where at least N_RTT_SAMPLE RTT samples have been obtained For rounds where at least N_RTT_SAMPLE RTT samples have been obtained
and currentRoundMinRTT and lastRoundMinRTT are valid, check if delay and currentRoundMinRTT and lastRoundMinRTT are valid, check to see if
increase triggers slow start exit: delay increase triggers slow start exit:
if ((rttSampleCount >= N_RTT_SAMPLE) AND if ((rttSampleCount >= N_RTT_SAMPLE) AND
(currentRoundMinRTT != infinity) AND (currentRoundMinRTT != infinity) AND
(lastRoundMinRTT != infinity)) (lastRoundMinRTT != infinity))
Compute a RTT Threshold clamped between MIN_RTT_THRESH and MAX_RTT_THRESH RttThresh = max(MIN_RTT_THRESH,
RttThresh = max(MIN_RTT_THRESH, min(lastRoundMinRTT / MIN_RTT_DIVISOR, MAX_RTT_THRESH)) min(lastRoundMinRTT / MIN_RTT_DIVISOR, MAX_RTT_THRESH))
if (currentRoundMinRTT >= (lastRoundMinRTT + RttThresh)) if (currentRoundMinRTT >= (lastRoundMinRTT + RttThresh))
cssBaselineMinRtt = currentRoundMinRTT cssBaselineMinRtt = currentRoundMinRTT
exit slow start and enter CSS exit slow start and enter CSS
For each arriving ACK in CSS, where N is the number of previously For each arriving ACK in CSS, where N is the number of previously
unacknowledged bytes acknowledged in the arriving ACK: unacknowledged bytes acknowledged in the arriving ACK:
Update the cwnd: Update the cwnd:
cwnd = cwnd + (min(N, L * SMSS) / CSS_GROWTH_DIVISOR) cwnd = cwnd + (min(N, L * SMSS) / CSS_GROWTH_DIVISOR)
Keep track of minimum observed RTT: Keep track of the minimum observed RTT:
currentRoundMinRTT = min(currentRoundMinRTT, currRTT) currentRoundMinRTT = min(currentRoundMinRTT, currRTT)
rttSampleCount += 1 rttSampleCount += 1
For CSS rounds where at least N_RTT_SAMPLE RTT samples have been For CSS rounds where at least N_RTT_SAMPLE RTT samples have been
obtained, check if current round's minRTT drops below baseline obtained, check to see if the current round's minRTT drops below
indicating that HyStart exit was spurious: baseline (cssBaselineMinRtt) indicating that slow start exit was
spurious:
if (currentRoundMinRTT < cssBaselineMinRtt) if (currentRoundMinRTT < cssBaselineMinRtt)
cssBaselineMinRtt = infinity cssBaselineMinRtt = infinity
resume slow start including HyStart++ resume slow start including HyStart++
CSS lasts at most CSS_ROUNDS rounds. If the transition into CSS CSS lasts at most CSS_ROUNDS rounds. If the transition into CSS
happens in the middle of a round, that partial round counts towards happens in the middle of a round, that partial round counts towards
the limit. the limit.
If CSS_ROUNDS rounds are complete, enter congestion avoidance by If CSS_ROUNDS rounds are complete, enter congestion avoidance by
setting ssthresh to current cwnd. setting the ssthresh to the current cwnd.
ssthresh = cwnd ssthresh = cwnd
If loss or ECN-marking is observed anytime during standard slow start If loss or Explicit Congestion Notification (ECN) marking is observed
or CSS, enter congestion avoidance by setting ssthresh to current at any time during standard slow start or CSS, enter congestion
cwnd. avoidance by setting the ssthresh to the current cwnd.
ssthresh = cwnd ssthresh = cwnd
4.3. Tuning constants and other considerations 4.3. Tuning Constants and Other Considerations
It is RECOMMENDED that a HyStart++ implementation use the following It is RECOMMENDED that a HyStart++ implementation use the following
constants: constants:
MIN_RTT_THRESH = 4 msec MIN_RTT_THRESH = 4 msec
MAX_RTT_THRESH = 16 msec MAX_RTT_THRESH = 16 msec
MIN_RTT_DIVISOR = 8 MIN_RTT_DIVISOR = 8
N_RTT_SAMPLE = 8 N_RTT_SAMPLE = 8
CSS_GROWTH_DIVISOR = 4 CSS_GROWTH_DIVISOR = 4
CSS_ROUNDS = 5 CSS_ROUNDS = 5
L = infinity if paced, L = 8 if non-paced L = infinity if paced, L = 8 if non-paced
These constants have been determined with lab measurements and real These constants have been determined with lab measurements and real-
world deployments. An implementation MAY tune them for different world deployments. An implementation MAY tune them for different
network characteristics. network characteristics.
The delay increase sensitivity is determined by MIN_RTT_THRESH and The delay increase sensitivity is determined by MIN_RTT_THRESH and
MAX_RTT_THRESH. Smaller values of MIN_RTT_THRESH may cause spurious MAX_RTT_THRESH. Smaller values of MIN_RTT_THRESH may cause spurious
exits from slow start. Larger values of MAX_RTT_THRESH may result in exits from slow start. Larger values of MAX_RTT_THRESH may result in
slow start not exiting until loss is encountered for connections on slow start not exiting until loss is encountered for connections on
large RTT paths. large RTT paths.
MIN_RTT_DIVISOR is a fraction of RTT to compute delay threshold. A MIN_RTT_DIVISOR is a fraction of RTT to compute the delay threshold.
smaller value would mean a bigger threshold and thus less sensitive A smaller value would mean a larger threshold and thus less
to delay increase, and vice versa. sensitivity to delay increase, and vice versa.
While all TCP implementations are REQUIRED to take at least one RTT While all TCP implementations are REQUIRED to take at least one RTT
sample each round, implementations of HyStart++ are RECOMMENDED to sample each round, implementations of HyStart++ are RECOMMENDED to
take at least N_RTT_SAMPLE RTT samples. Using lower values of take at least N_RTT_SAMPLE RTT samples. Using lower values of
N_RTT_SAMPLE will lower the accuracy of the measured RTT for the N_RTT_SAMPLE will lower the accuracy of the measured RTT for the
round; higher values will improve accuracy at the cost of more round; higher values will improve accuracy at the cost of more
processing. processing.
The minimum value of CSS_GROWTH_DIVISOR MUST be at least 2. A value The minimum value of CSS_GROWTH_DIVISOR MUST be at least 2. A value
of 1 results in the same aggressive behavior as regular slow start. of 1 results in the same aggressive behavior as regular slow start.
Values larger than 4 will cause the algorithm to be less aggressive Values larger than 4 will cause the algorithm to be less aggressive
and maybe less performant. and maybe less performant.
Smaller values of CSS_ROUNDS may miss detecting jitter and larger Smaller values of CSS_ROUNDS may miss detecting jitter, and larger
values may limit performance. values may limit performance.
Packet pacing [ASA00] is a possible mechanism to avoid large bursts Packet pacing [ASA00] is a possible mechanism to avoid large bursts
and their associated harm. A paced TCP implementation SHOULD use L = and their associated harm. A paced TCP implementation SHOULD use L =
infinity. Burst concerns are mitigated by pacing and this setting infinity. Burst concerns are mitigated by pacing, and this setting
allows for optimal cwnd growth on modern networks. allows for optimal cwnd growth on modern networks.
For TCP implementations that pace to mitigate burst concerns, L For TCP implementations that pace to mitigate burst concerns, L
values smaller than INFINITY may suffer performance problems due to values smaller than infinity may suffer performance problems due to
slow cwnd growth in high speed networks. For non-paced TCP slow cwnd growth in high-speed networks. For non-paced TCP
implementations, L values smaller than 8 may suffer performance implementations, L values smaller than 8 may suffer performance
problems due to slow cwnd growth in high speed networks; L values problems due to slow cwnd growth in high-speed networks; L values
larger than 8 may cause an increase in burstiness and thereby loss larger than 8 may cause an increase in burstiness and thereby loss
rates, and result in poor performance. rates, and result in poor performance.
An implementation SHOULD use HyStart++ only for the initial slow An implementation SHOULD use HyStart++ only for the initial slow
start (when ssthresh is at its initial value of arbitrarily high per start (when the ssthresh is at its initial value of arbitrarily high
[RFC5681]) and fall back to using traditional slow start for the per [RFC5681]) and fall back to using standard slow start for the
remainder of the connection lifetime. This is acceptable because remainder of the connection lifetime. This is acceptable because
subsequent slow starts will use the discovered ssthresh value to exit subsequent slow starts will use the discovered ssthresh value to exit
slow start and avoid the overshoot problem. An implementation MAY slow start and avoid the overshoot problem. An implementation MAY
use HyStart++ to grow the restart window ([RFC5681]) after a long use HyStart++ to grow the restart window [RFC5681] after a long idle
idle period. period.
In application limited scenarios, the amount of data in flight could In application-limited scenarios, the amount of data in flight could
fall below the bandwidth-delay product (BDP) and result in smaller fall below the bandwidth-delay product (BDP) and result in smaller
RTT samples which can trigger an exit back to slow start. It is RTT samples, which can trigger an exit back to slow start. It is
expected that a connection might oscillate between CSS and slow start expected that a connection might oscillate between CSS and slow start
in such scenarios. But this behavior will neither result in a in such scenarios. But this behavior will neither result in a
connection prematurely entering congestion avoidance nor cause connection prematurely entering congestion avoidance nor cause
overshooting compared to slow start. overshooting compared to slow start.
5. Deployments and Performance Evaluations 5. Deployments and Performance Evaluations
As of February 2023, HyStart++ as described in this document has been At the time of this writing, HyStart++ as described in this document
default enabled for all TCP connections in the Windows operating has been default enabled for all TCP connections in the Windows
system for over two years with pacing disabled and an actual L = 8. operating system for over two years with pacing disabled and an
actual L = 8.
In lab measurements with Windows TCP, HyStart++ shows both goodput In lab measurements with Windows TCP, HyStart++ shows goodput
improvements as well as reductions in packet loss and retransmissions improvements as well as reductions in packet loss and retransmissions
compared to traditional slow start. For example, across a variety of compared to standard slow start. For example, across a variety of
tests on a 100 Mbps link with a bottleneck buffer size of bandwidth- tests on a 100 Mbps link with a bottleneck buffer size of bandwidth-
delay product, HyStart++ reduces bytes retransmitted by 50% and delay product, HyStart++ reduces bytes retransmitted by 50% and
retransmission timeouts (RTOs) by 36%. retransmission timeouts (RTOs) by 36%.
In an A/B test where we compare HyStart++ draft 01 to traditional In an A/B test where we compared an implementation of HyStart++
slow start across a large Windows device population, out of 52 (based on an earlier draft version of this document) to standard slow
billion TCP connections, 0.7% of connections move from 1 RTO to 0 start across a large Windows device population, out of 52 billion TCP
RTOs and another 0.7% connections move from 2 RTOs to 1 RTO with connections, 0.7% of connections move from 1 RTO to 0 RTOs and
HyStart++. This test did not focus on send-heavy connections and the another 0.7% of connections move from 2 RTOs to 1 RTO with HyStart++.
impact on send-heavy connections is likely much higher. We plan to This test did not focus on send-heavy connections, and the impact on
conduct more such production experiments to gather more data in the send-heavy connections is likely much higher. We plan to conduct
future. more such production experiments to gather more data in the future.
6. Security Considerations 6. Security Considerations
HyStart++ enhances slow start and inherits the general security HyStart++ enhances slow start and inherits the general security
considerations discussed in [RFC5681]. considerations discussed in [RFC5681].
An attacker can cause Hystart++ to exit slow start prematurely and An attacker can cause HyStart++ to exit slow start prematurely and
impair the performance of a TCP connection by, for example, dropping impair the performance of a TCP connection by, for example, dropping
data packets or their acknowledgements. data packets or their acknowledgments.
The ACK division attack outlined in [SCWA99] does not affect The ACK division attack outlined in [SCWA99] does not affect
Hystart++ because the congestion window increase in Hystart++ is HyStart++ because the congestion window increase in HyStart++ is
based on the number of bytes newly acknowledged in each arriving ACK based on the number of bytes newly acknowledged in each arriving ACK
rather than by a particular constant on each arriving ACK. rather than by a particular constant on each arriving ACK.
7. IANA Considerations 7. IANA Considerations
This document has no actions for IANA. This document has no IANA actions.
8. Acknowledgements
During the discussions of this work on the TCPM mailing list, in
working group meetings, helpful comments, critiques, and reviews were
received from (listed alphabetically by last name): Mark Allman, Bob
Briscoe, Neal Cardwell, Yuchung Cheng, Junho Choi, Martin Duke, Reese
Enghardt, Christian Huitema, Ilpo Järvinen, Yoshifumi Nishida,
Randall Stewart, and Michael Tuexen.
9. References 8. References
9.1. Normative References 8.1. Normative References
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, Requirement Levels", BCP 14, RFC 2119,
DOI 10.17487/RFC2119, March 1997, DOI 10.17487/RFC2119, March 1997,
<https://www.rfc-editor.org/info/rfc2119>. <https://www.rfc-editor.org/info/rfc2119>.
[RFC5681] Allman, M., Paxson, V., and E. Blanton, "TCP Congestion [RFC5681] Allman, M., Paxson, V., and E. Blanton, "TCP Congestion
Control", RFC 5681, DOI 10.17487/RFC5681, September 2009, Control", RFC 5681, DOI 10.17487/RFC5681, September 2009,
<https://www.rfc-editor.org/info/rfc5681>. <https://www.rfc-editor.org/info/rfc5681>.
[RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
May 2017, <https://www.rfc-editor.org/info/rfc8174>. May 2017, <https://www.rfc-editor.org/info/rfc8174>.
9.2. Informative References 8.2. Informative References
[ASA00] Aggarwal, A., Savage, S., and T. Anderson, "Understanding [ASA00] Aggarwal, A., Savage, S., and T. Anderson, "Understanding
the Performance of TCP Pacing", Proceedings IEEE INFOCOM the performance of TCP pacing", Proceedings IEEE INFOCOM
2000, DOI 10.1109/INFCOM.2000.832483, 2000, 2000, DOI 10.1109/INFCOM.2000.832483, March 2000,
<https://doi.org/10.1109/INFCOM.2000.832483>. <https://doi.org/10.1109/INFCOM.2000.832483>.
[HyStart] Ha, S. and I. Ree, "Taming the elephants: New TCP slow [HyStart] Ha, S. and I. Rhee, "Taming the elephants: New TCP slow
start", Computer Networks vol. 55, no. 9, pp. 2092-2110, start", Computer Networks vol. 55, no. 9, pp. 2092-2110,
DOI 10.1016/j.comnet.2011.01.014, 2011, DOI 10.1016/j.comnet.2011.01.014, June 2011,
<https://doi.org/10.1016/j.comnet.2011.01.014>. <https://doi.org/10.1016/j.comnet.2011.01.014>.
[RFC1122] Braden, R., Ed., "Requirements for Internet Hosts - [RFC1122] Braden, R., Ed., "Requirements for Internet Hosts -
Communication Layers", STD 3, RFC 1122, Communication Layers", STD 3, RFC 1122,
DOI 10.17487/RFC1122, October 1989, DOI 10.17487/RFC1122, October 1989,
<https://www.rfc-editor.org/info/rfc1122>. <https://www.rfc-editor.org/info/rfc1122>.
[RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191,
DOI 10.17487/RFC1191, November 1990, DOI 10.17487/RFC1191, November 1990,
<https://www.rfc-editor.org/info/rfc1191>. <https://www.rfc-editor.org/info/rfc1191>.
skipping to change at page 9, line 37 skipping to change at line 411
[RFC9002] Iyengar, J., Ed. and I. Swett, Ed., "QUIC Loss Detection [RFC9002] Iyengar, J., Ed. and I. Swett, Ed., "QUIC Loss Detection
and Congestion Control", RFC 9002, DOI 10.17487/RFC9002, and Congestion Control", RFC 9002, DOI 10.17487/RFC9002,
May 2021, <https://www.rfc-editor.org/info/rfc9002>. May 2021, <https://www.rfc-editor.org/info/rfc9002>.
[RFC9260] Stewart, R., Tüxen, M., and K. Nielsen, "Stream Control [RFC9260] Stewart, R., Tüxen, M., and K. Nielsen, "Stream Control
Transmission Protocol", RFC 9260, DOI 10.17487/RFC9260, Transmission Protocol", RFC 9260, DOI 10.17487/RFC9260,
June 2022, <https://www.rfc-editor.org/info/rfc9260>. June 2022, <https://www.rfc-editor.org/info/rfc9260>.
[SCWA99] Savage, S., Cardwell, N., Wetherall, D., and T. Anderson, [SCWA99] Savage, S., Cardwell, N., Wetherall, D., and T. Anderson,
"TCP congestion control with a misbehaving receiver", ACM "TCP congestion control with a misbehaving receiver", ACM
Computer Communication Review, 29(5), SIGCOMM Computer Communication Review, vol. 29, issue 5,
DOI 10.1145/505696.505704, 1999, pp. 71-78, DOI 10.1145/505696.505704, October 1999,
<https://doi.org/10.1145/505696.505704>. <https://doi.org/10.1145/505696.505704>.
Acknowledgments
During the discussions of this work on the TCPM mailing list and in
working group meetings, helpful comments, critiques, and reviews were
received from (listed alphabetically by last name) Mark Allman, Bob
Briscoe, Neal Cardwell, Yuchung Cheng, Junho Choi, Martin Duke, Reese
Enghardt, Christian Huitema, Ilpo Järvinen, Yoshifumi Nishida,
Randall Stewart, and Michael Tüxen.
Authors' Addresses Authors' Addresses
Praveen Balasubramanian Praveen Balasubramanian
Confluent Confluent
899 West Evelyn Ave 899 West Evelyn Ave
Mountain View, CA 94041 Mountain View, CA 94041
United States of America United States of America
Email: pravb.ietf@gmail.com Email: pravb.ietf@gmail.com
Yi Huang Yi Huang
Microsoft Microsoft
One Microsoft Way One Microsoft Way
Redmond, WA 94052 Redmond, WA 98052
United States of America United States of America
Phone: +1 425 703 0447 Phone: +1 425 703 0447
Email: huanyi@microsoft.com Email: huanyi@microsoft.com
Matt Olson Matt Olson
Microsoft Microsoft
One Microsoft Way
Redmond, WA 98052
United States of America
Phone: +1 425 538 8598 Phone: +1 425 538 8598
Email: maolson@microsoft.com Email: maolson@microsoft.com
 End of changes. 59 change blocks. 
154 lines changed or deleted 165 lines changed or added

This html diff was produced by rfcdiff 1.48.