Internet Engineering Task Force Yogesh Swami INTERNET DRAFT Khiem Le File: draft-swami-tsvwg-tcp-dclor-01.txt Nokia Research Center Dallas Apr 2003 Expires: Oct 2003 DCLOR: De-correlated Loss Recovery using SACK option for spurious timeouts. Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of [RFC2026]. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at 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 Abstract A spurious timeout in TCP forces the sender to unnecessarily retransmit one complete congestion window of data into the network. In addition, TCP uses the rate of arrival of ACKs as the basic criterion for congestion control. TCP makes the assumption that the rate at which ACKs are received reflects the end-to-end state of the network in terms of congestion. But after a spurious-timeout, the ACKs don't reflect the end-to-end congestion state of the network, but only a part of it. In these cases, the slow-start behavior after a timeout can further add to network congestion. In this draft we propose changes to the TCP sender (no change is needed for TCP receiver) that can be used to solve the problem of both redundant- retransmission and network congestion after a spurious timeout. Expires: Oct 2003 [Page 1] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 1. Introduction The response of a TCP sender after a retransmission timeout is governed by the underlying assumption that a mid-stream timeout can occur only if there is heavy congestion--manifested as packet loss--in the network. Even though loss is often caused by congestion, the loss recovery algorithm itself should only answer the question of "what" data (i.e., what sequence number of data ) to send. While on the other hand, the congestion control algorithm should answer the question of "how much" data to send. But after a timeout, TCP addresses the issues of loss recovery and congestion control using a single mechanism--send one segment per round trip timeout (RTO) (answers the "how much" question) until an acknowledgment is received. The single segment sent is always the first unacknowledged outstanding packet in the retransmission queue (answers the "what" question). Since the present TCP's loss recovery and congestion control algorithms are coupled together, we call this "Correlated Loss Recovery (CLOR)." Although the assumption that a timeout can occur only if there is severe congestion is valid for traditional wire-line networks, it does not hold good for some other types of networks--networks where packets can be stalled "in the network" for a significant duration without being discarded. Typical examples of such networks are cellular networks. In cellular networks, the link layer can experience a relatively long disruption due to errors, and the link layer protocol can keep these packets-in-error buffered as long as the link layer disruption lasts. In this document we present an alternative approach to loss recovery and congestion control that "De-Correlates" Loss Recovery from congestion congestion and allows independent choice on using a particular TCP sequence number without compromising on the congestion control principles of [RFC2581][RFC2914][RFC2861]. Although several drafts [LM02][LG03][SK03][BA02] have been presented on this topic, we believe that none of them fully considers all the problems associated with spurious timeouts. In the following section we first describe these problems in more detail and then describe the DCLOR mechanism in section-3. 2. Problem Description. Let us assume that a TCP sender has sent N packets, p(1) ... p(N), into the network and it's waiting for the ACK of p(1) (Figure-1). Due to bad network conditions or some other problem, these packets are excessively delayed at some some intermediary node NDN. Unlike standard IP routers, the NDN keeps these packets buffered for a Expires: Oct 2003 [Page 2] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 relatively long period of time until these packets are forwarded to their intended recipient. This excessive delay forces the TCP sender to timeout and enter slow start. Figure-1 TCP-Sender NDN TCP-Receiver ..... |----p(1)------>| | ^ |----p(2)------>| | : | . | | RTT=D | . | | : | . | | ..... |----p(N)------>| | | ^ | | | : | | | RTO | | | : | | | V |----p(1)-->| ... |----p1(1)----->|<---a(1)---|... L | | | ... |<----a(1)------|----p(2)-->| |->p1(2),p1(3)->|<---a(2)---|... | . | . | | . | . | | . | . | | |<---a(N)---| | |---p1(1)-->| | |<---a(N)---| | | | As far as the sender is concerned, a timeout is always interpreted as heavy congestion. The TCP sender therefore makes the assumption that all packets between p(1) and p(N) were lost in the network. To recover from this misconstrued loss, the TCP sender retransmits P1(1) ( Px(k) represents the xth retransmission of packet with sequence number k), and waits for the ACK a(1). After some period of time when the network conditions at NDN improve, the queued in packets are finally dispatched to their intended recipient; in response the TCP receiver generates the ACK a(1). When the TCP sender receives a(1), it's fooled into believing that a(1) was generated in response to the retransmitted packet p1(1), while in reality a(1) was generated in response to the originally transmitted packet p(1). When the sender receives a(1), it increases its congestion window to two, and retransmits p1(2) and p1(3). As the sender receives more acknowledgments, it continues with Expires: Oct 2003 [Page 3] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 retransmissions and finally starts sending new data. The following two sub sections examine the problems associated with the above-mentioned TCP behavior. 2.1 Redundant Data Retransmission The obvious and relatively easy-to-solve inefficiency of the above algorithm is that the entire congestion window worth of data is unnecessarily retransmitted. Although such retransmissions are harmless to high-bandwidth, well-provisioned, backbone links (so long they are infrequent), it could severely degrade the performance of slow links. In cases where bandwidth is a commodity at a premium, (e.g., cellular networks), unnecessary retransmission can also be costly. 2.2 Congestion after Spurious Timeout To analyze network congestion after spurious timeout, we compute the worst case scenario packet loss in the system--assuming only TCP connections to be present. After the spurious timeout, the TCP sender sets its SS_THRESH to N/2. Therefore, for the first N/2 ACKs received (i.e., ACK a(1) to a(N/2) ), the TCP sender will grow its congestion window by one and reach the SS_THRESH value of N/2. For each ACK received, the TCP sender sends 2 packets. Therefore, by the end of the slow start, the TCP sender would have sent 2*(N/2) packets into the network. For the remaining N/2 ACKs (i.e., ACKs between a(N/2+1) to a(N)) the TCP sender will remain in the congestion avoidance phase and send one packet for each ACK received--sending N/2 more data segments. The net amount of data sent is therefore N/2 + N = 3N/2. Please note that the entire 3N/2 packets are injected into the network within a time period less than or equal to RTT in most cases. The number of data segments that left the network during this time is only N. Therefore, N/2 packets out of 3N/2 packets will be lost with a very high probability. These N/2 lost packets, however, need not come from the same connection, and such a data-burst will unnecessarily penalize all the competing TCP connections that share the same bottleneck router. Going further ahead, let us assume there are M competing TCP connections that share the same bottleneck router(s) with C(0)(Figure-2). During the period of time while C(0) is stalled, the TCP sender of C(0) does not use its network resources--the buffer space--on the bottleneck router(s). The competing connections, Expires: Oct 2003 [Page 4] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 C(1)... C(M), however see this lack of activity as resource availability and start growing their window by at least one segment per RTT during this time period (by virtue of linear window increase during congestion avoidance phase). For simplicity reasons, we assume that each of these connections has the same round trip time of RTT, and the idle time for C(0) is k*RTT (where k > RTO/RTT). Under these assumptions, each of these competing connections will increase their congestion window by k segments. Therefore the amount of packets lost in the network due to slow start can be as high as: N/2 + M*k ... (4) the first term in the above equation is the packet loss due to slow start, while the second term is the loss due to window growth of completing connections (if the competing connections were in slow start the response could have been worse). Figure-2 C(1) C(2)... C(M) | | ... | | | ... | | | ... | V V ... V \ \ / \ \ / \ \ / +------X--X--X---+ +------------------+ Defaulting | | | | C(0) ----------->| Bottleneck |------>|Buffered packets |---> connection | router | | | +-----X--X----X--+ +------------------+ | | | | | | c(1)c(2) C(M) Based on the above equation, we note that the congestion state of the network depends upon the duration of spurious timeout. In our reponse algorithm we therefore take the time duration of spurious timeout into account reduce the data rate by half every RTO. Please note that this scheme works well only when the number of competing connections M does not vary too much while C(0) was stalled. A more conservative response algorithm should reduce the data rate to INIT_WINDOW if M is not bounded. In the following sections we describe an algorithm that solves the problem of both redundant retransmission and packet loss after a spurious timeout. Expires: Oct 2003 [Page 5] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 3. De-correlated Loss Recovery (DCLOR) The basic idea behind DCLOR is to send a new data segment from outside the sender's retransmission queue and wait for the ACK or SACK of the new data before initiating the response algorithm. Unlike slow-start where the response algorithm starts immediately after receiving the first ACK, DCLOR waits for the ACK/SACK of the new data sent after timeout before initiating loss recovery. The SACK block for new data contains sufficient information to determine all the packets that were lost into the network. Once the sequence number of lost packets is determined, the TCP sender grows its congestion window as determined by the SS_THRESH and it's congestion window. 3.1 Probe phase after a timeout The following steps describe the response of a TCP sender on a timeout: 1. If the timeout occurs before the 3 way handshake is complete, the TCP sender's behavior is unchanged, 2. After each timeout, the TCP sender MUST set its congestion window to: cwnd = max( cwnd >> 1, IINIT_WINDOW). The value of SS_THRESH MUST be left UNCHANGED at this point. The TCP sender should also count the number of packets in flight at this time, and keep it in a state variable stale_outstanding. 3. The TCP sender SHOULD also reset all the SACK tag bits in its retransmission queue if this the first timeout. 4. Instead of sending the first unacknowledged packet P1 after a timeout, the TCP sender should *disregard* its congestion window and send ONE NEW MSS size data Pn+1. The TCP sender should also store the sequence number of the new segment in a new state variable called SS_PTR (for slow start pointer). If the sender does not have any new data outside its retransmission queue, or if the receiver's flow control window cannot sustain any new data, the TCP sender SHOULD send the highest sequence numbered MSS sized data chunk from its retransmission queue (i.e., it should send the last packet from its retransmission queue). Expires: Oct 2003 [Page 6] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 5. A TCP sender MUST repeat step-2 to step-4 until it enters the Timeout-Recovery state as described in step 6. 3.2 Congestion Control After the probe phase 6. For each ACK received with the ACK-sequence number less than SS_PTR, regardless of the value of the SS_THRESH, the TCP sender SHOULD NOT grow it's congestion window. If the ACK contains a new SACK block, the SACK tag SHOULD be set in the corresponding data packet. If new segments were ACKed, and the congestion window allows, the TCP sender SHOULD send new data. (Note: the idea here is that the congestion window should not be grown in response to stale ACKs since these ACKs don't reflect the end to end state of the network). In addition, the TCP sender SHOULD NOT take any timer sample for the stale ACKs. (NOTE: We do not attempt to change the RTT calculation in an ad-hoc manner; we believe that this is a reaseach problem that needs better network modelling before an appropriate timer calculation can be found) 7. Step-6 continues until the TCP sender receives an ACK acking a sequence number greater than SS_PTR, or it receives a SACK block covering the sequence number greater than SS_PTR. If the sender receives a SACK block containing SS_PTR, i.e., if there is a packet loss in the stalled window, it SHOULD go to step-8. If the sender receives an ACK that acknowledges SS_PTR, i.e., if no packets were lost from the stalled window, it SHOULD go to step-10. NOTE: In our previous experiments we had set the congestion window to one MSS after a spurious timeout, however this algorithm prerforms better if there is moderate load on the routers and the number of competing connections do not vary a lot duing the stalling period. In case of heavy load, setting the congestion window to INIT_WINDOW still performs better. We believe that using the present congestion response make a fair compromise for different scenarios. 3.3 Timeout-Recovery: recovering lost packets after timeout 8. The TCP sender traverses the retransmission queue and marks all the packets without any SACK tag as lost. The TCP sender also updates its packets-in-flight (pipe) based on the SACK tags and the lost segment information (the packets-in-flight (pipe) should be ZERO after the update). Expires: Oct 2003 [Page 7] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 Please note that unlike Fast-Retransmit and Fast-recovery, DCLOR uses only one SACK block containing SS_PTR to mark packets as lost. This is because we do not expect packet reordering to exist over the period of RTO. 9. The TCP sender should update its SS_THRESH, as: SS_THRESH= stale_outstanding >> 1 (step-2) 10. The TCP sender SHOULD set its congestion window to cwnd+1. If packets were lost into the network (i.e., if a SACK for SS_PTR was received), the TCP sender should start by sending packets with lowest sequence number; else it should continue with new data. (Note: for each new SACK block received, the sender should send a segment--lost or new--and therefore the problem of duplicate ACKs is not of concern here.) The sender should follow the normal window growth strategy based on the value of SS_THRESH after this step. Please note that with a pure ACK acknowledging SS_PTR, the TCP sender does not update the SS_THRESH value (it directly enters step-10 from step-7). This prevents a TCP sender from setting its SS_THRESH to a very small values if the spurious timeout occurs at the start of the connection. 4. Data Delivery To Upper Layers If a TCP sender loses its entire congestion window worth of data, sending new data after timeout prevents a TCP receiver from forwarding the new data to the upper layers immediately. However, once the SACK for this new data is received, the TCP sender will send the first lost segment. This essentially means that data delivery to the upper layers could be delayed by at most one RTT when all the packets are lost in the network. This, however, does not affect the throughput of the connection in any way. If a timeout has occurred, then the data delivery to the upper layers has already been excessively delayed. Delaying it by another round trip is not a serious problem. Please note that reliability and timeliness are two conflicting issues and one cannot gain on one without sacrificing something else on the other. 5. Security Considerations The TCP SACK information is meant to be advisory, and a TCP receiver is allowed--though strongly discouraged--to discard data blocks the receiver has already SACKed [RFC2018]. Please note however that even Expires: Oct 2003 [Page 8] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 if the TCP sender discards the data block it received, it MUST still send the SACK block for at least the recent most data received. Therefore in spite of SACK reneging, DCLOR will work without any deadlocks. A SACK implementation is also allowed not to send a SACK block even though the TCP sender and receiver might have agreed to SACK- Permitted option at the start of the connection. In these cases, however, if the receiver sends one SACK block, it must send SACK blocks for the rest of the connection. Because of the above mentioned leniency in implementation, its possible that a TCP receiver may agree on SACK-Permitted option, and yet not send any SACK blocks. To make DCLOR robust under these circumstances, DCLOR SHOULD NOT be invoked unless the sender has seen at least one SACK block before timeout. We, however, believe that once the SACK-Permitted option is accepted, the TCP sender MUST send a SACK block--even though that block might finally be discarded. Otherwise, the SACK-Permitted option is completely redundant and serves little purpose. To the best of our knowledge, almost all SACK implementations send a SACK block if they have accepted the SACK-Permitted option. 6. References [RFC2581] M. Allman, V. Paxson, W. Stevens. "TCP Congestion Control," Apr, 1999. [RFC2914] S. Floyd, "Congestion Control Principles," Sep 2002. [RFC2861] M. Handley, J. Padhye, S. Floyd. "TCP Congestion Window Validation," Jun 2000. [BAFW03] E. Blanton, M. Allman, K. Fall, L. Wang, "Conservative SACK-based Loss Recovery Algorithm for TCP," draft- allman-tcp-sack-13.txt. Internet draft; work in progress. Oct 2002. [RFC2018] M. Mathis, J. Mahdavi, S. Floyd, A. Romanow, "TCP Selective Acknowledgment Options," Oct 1996. [RFC2883] S. Floyd, J. Mahdavi, M. Mathis, M. Podolsky, "An Extension to the Selective Acknowledgment (SACK) Option for TCP," Jul 2000. [LM02] R. Ludwig, M. Meyer. "The Eiffel Detection Algorithm for TCP." Internet draft; work in progress, draft-ietf- tsvwg-tcp-eifel-alg-07.txt, Dec 2002. Expires: Oct 2003 [Page 9] draft-swami-tsvwg-tcp-dclor-01.txt Apr 2003 [LG03] R. Ludwig, A. Gurtov, "The Eifel Response Algorithm for TCP." Internet draft; work in progress, draft-ietf-tsvwg- tcp-eifel-response-03.txt, Mar 2003. [SK03] P. Sarolahti, M. Kojo. "F-RTO: A TCP RTO Recovery Algorithm for Avoiding Unnecessary Retransmissions." Internet draft; work in progress. draft-sarolahti-tsvwg- tcp-frto-03.txt, Jan 2003. [RFC2988] V. Paxon, M. Allman. "Computing TCP's Retransmission Timer," Nov 2000. [BA02] E. Blanton, M. Allman, "Using TCP DSACKs and SCTP Duplicate TSNs to Detect Spurious Retransmissions," Internet draft; work in progress, draft-blanton-dsack- use-02.txt, Oct 2002. 7. IPR Statement The IETF has been notified of intellectual property rights claimed in regard to some or all of the specification contained in this document. For more information consult the on-line list of claimed rights at http://www.ietf.org/ipr. Author's Address: Yogesh Prem Swami Khiem Le Nokia Research Center Nokia Research Center 6000 Connection Drive 6000 Connection Drive Irving TX-75063 Irving TX-75063 USA USA Phone: +1 972-374-0669 Phone: +1 972-894-4882 Email: yogesh.swami@nokia.com Email: khiem.le@nokia.com Expires: Oct 2003 [Page 10]