Fastly

jri.ietf@gmail.com

Google

ianswett@google.com

Transport QUIC This document describes loss detection and congestion control mechanisms for QUIC. Discussion of this draft takes place on the QUIC working group mailing list (quic@ietf.org), which is archived at https://mailarchive.ietf.org/arch/search/?email_list=quic. Working Group information can be found at https://github.com/quicwg; source code and issues list for this draft can be found at https://github.com/quicwg/base-drafts/labels/-recovery.

QUIC is a new multiplexed and secure transport atop UDP. QUIC builds on decades of transport and security experience, and implements mechanisms that make it attractive as a modern general-purpose transport. The QUIC protocol is described in . QUIC implements the spirit of existing TCP loss recovery mechanisms, described in RFCs, various Internet-drafts, and also those prevalent in the Linux TCP implementation. This document describes QUIC congestion control and loss recovery, and where applicable, attributes the TCP equivalent in RFCs, Internet-drafts, academic papers, and/or TCP implementations.

The key words “MUST”, “MUST NOT”, “REQUIRED”, “SHALL”, “SHALL NOT”, “SHOULD”, “SHOULD NOT”, “RECOMMENDED”, “NOT RECOMMENDED”, “MAY”, and “OPTIONAL” in this document are to be interpreted as described in BCP 14 when, and only when, they appear in all capitals, as shown here. Definitions of terms that are used in this document: Any packet containing only one or more ACK frame(s). Packets are considered in-flight when they have been sent and neither acknowledged nor declared lost, and they are not ACK-only. All frames besides ACK or PADDING are considered ack-eliciting. Packets that contain ack-eliciting frames elicit an ACK from the receiver within the maximum ack delay and are called ack-eliciting packets. Packets containing CRYPTO data sent in Initial or Handshake packets. Packets that do not increase the largest received packet number for its packet number space by exactly one. Packets arrive out of order when earlier packets are lost or delayed.

All transmissions in QUIC are sent with a packet-level header, which indicates the encryption level and includes a packet sequence number (referred to below as a packet number). The encryption level indicates the packet number space, as described in . Packet numbers never repeat within a packet number space for the lifetime of a connection. Packet numbers monotonically increase within a space, preventing ambiguity. This design obviates the need for disambiguating between transmissions and retransmissions and eliminates significant complexity from QUIC’s interpretation of TCP loss detection mechanisms. QUIC packets can contain multiple frames of different types. The recovery mechanisms ensure that data and frames that need reliable delivery are acknowledged or declared lost and sent in new packets as necessary. The types of frames contained in a packet affect recovery and congestion control logic: All packets are acknowledged, though packets that contain no ack-eliciting frames are only acknowledged along with ack-eliciting packets. Long header packets that contain CRYPTO frames are critical to the performance of the QUIC handshake and use shorter timers for acknowledgement and retransmission. Packets that contain only ACK frames do not count toward congestion control limits and are not considered in-flight. PADDING frames cause packets to contribute toward bytes in flight without directly causing an acknowledgment to be sent.

Readers familiar with TCP’s loss detection and congestion control will find algorithms here that parallel well-known TCP ones. Protocol differences between QUIC and TCP however contribute to algorithmic differences. We briefly describe these protocol differences below.

QUIC uses separate packet number spaces for each encryption level, except 0-RTT and all generations of 1-RTT keys use the same packet number space. Separate packet number spaces ensures acknowledgement of packets sent with one level of encryption will not cause spurious retransmission of packets sent with a different encryption level. Congestion control and round-trip time (RTT) measurement are unified across packet number spaces.

TCP conflates transmission order at the sender with delivery order at the receiver, which results in retransmissions of the same data carrying the same sequence number, and consequently leads to “retransmission ambiguity”. QUIC separates the two: QUIC uses a packet number to indicate transmission order, and any application data is sent in one or more streams, with delivery order determined by stream offsets encoded within STREAM frames. QUIC’s packet number is strictly increasing within a packet number space, and directly encodes transmission order. A higher packet number signifies that the packet was sent later, and a lower packet number signifies that the packet was sent earlier. When a packet containing ack-eliciting frames is detected lost, QUIC rebundles necessary frames in a new packet with a new packet number, removing ambiguity about which packet is acknowledged when an ACK is received. Consequently, more accurate RTT measurements can be made, spurious retransmissions are trivially detected, and mechanisms such as Fast Retransmit can be applied universally, based only on packet number. This design point significantly simplifies loss detection mechanisms for QUIC. Most TCP mechanisms implicitly attempt to infer transmission ordering based on TCP sequence numbers - a non-trivial task, especially when TCP timestamps are not available.

QUIC ACKs contain information that is similar to TCP SACK, but QUIC does not allow any acked packet to be reneged, greatly simplifying implementations on both sides and reducing memory pressure on the sender.

QUIC supports many ACK ranges, opposed to TCP’s 3 SACK ranges. In high loss environments, this speeds recovery, reduces spurious retransmits, and ensures forward progress without relying on timeouts.

QUIC ACKs explicitly encode the delay incurred at the receiver between when a packet is received and when the corresponding ACK is sent. This allows the receiver of the ACK to adjust for receiver delays, specifically the delayed ack timer, when estimating the path RTT. This mechanism also allows a receiver to measure and report the delay from when a packet was received by the OS kernel, which is useful in receivers which may incur delays such as context-switch latency before a userspace QUIC receiver processes a received packet.

QUIC SHOULD delay sending acknowledgements in response to packets, but MUST NOT excessively delay acknowledgements of ack-eliciting packets. Specifically, implementations MUST attempt to enforce a maximum ack delay to avoid causing the peer spurious timeouts. The maximum ack delay is communicated in the max_ack_delay transport parameter and the default value is 25ms. An acknowledgement SHOULD be sent immediately upon receipt of a second ack-eliciting packet. QUIC recovery algorithms do not assume the peer sends an ACK immediately when receiving a second ack-eliciting packet. In order to accelerate loss recovery and reduce timeouts, the receiver SHOULD send an immediate ACK after it receives an out-of-order packet. It could send immediate ACKs for in-order packets for a period of time that SHOULD NOT exceed 1/8 RTT unless more out-of-order packets arrive. If every packet arrives out-of- order, then an immediate ACK SHOULD be sent for every received packet. Similarly, packets marked with the ECN Congestion Experienced (CE) codepoint in the IP header SHOULD be acknowledged immediately, to reduce the peer’s response time to congestion events. As an optimization, a receiver MAY process multiple packets before sending any ACK frames in response. In this case the receiver can determine whether an immediate or delayed acknowledgement should be generated after processing incoming packets.

In order to quickly complete the handshake and avoid spurious retransmissions due to crypto retransmission timeouts, crypto packets SHOULD use a very short ack delay, such as the local timer granularity. ACK frames MAY be sent immediately when the crypto stack indicates all data for that packet number space has been received.

When an ACK frame is sent, one or more ranges of acknowledged packets are included. Including older packets reduces the chance of spurious retransmits caused by losing previously sent ACK frames, at the cost of larger ACK frames. ACK frames SHOULD always acknowledge the most recently received packets, and the more out-of-order the packets are, the more important it is to send an updated ACK frame quickly, to prevent the peer from declaring a packet as lost and spuriously retransmitting the frames it contains. Below is one recommended approach for determining what packets to include in an ACK frame.

When a packet containing an ACK frame is sent, the largest acknowledged in that frame may be saved. When a packet containing an ACK frame is acknowledged, the receiver can stop acknowledging packets less than or equal to the largest acknowledged in the sent ACK frame. In cases without ACK frame loss, this algorithm allows for a minimum of 1 RTT of reordering. In cases with ACK frame loss and reordering, this approach does not guarantee that every acknowledgement is seen by the sender before it is no longer included in the ACK frame. Packets could be received out of order and all subsequent ACK frames containing them could be lost. In this case, the loss recovery algorithm may cause spurious retransmits, but the sender will continue making forward progress.

Round-trip time (RTT) is calculated when an ACK frame arrives by computing the difference between the current time and the time the largest acked packet was sent. An RTT sample MUST NOT be taken for a packet that is not newly acknowledged or not ack-eliciting. When RTT is calculated, the ack delay field from the ACK frame SHOULD be limited to the max_ack_delay specified by the peer. Limiting ack_delay to max_ack_delay ensures a peer specifying an extremely small max_ack_delay doesn’t cause more spurious timeouts than a peer that correctly specifies max_ack_delay. It SHOULD be subtracted from the RTT as long as the result is larger than the min_rtt. If the result is smaller than the min_rtt, the RTT should be used, but the ack delay field should be ignored. A sender calculates both smoothed RTT (SRTT) and RTT variance (RTTVAR) similar to those specified in , see . A sender takes an RTT sample when an ACK frame is received that acknowledges a larger packet number than before (see ). A sender will take multiple RTT samples per RTT when multiple such ACK frames are received within an RTT. When multiple samples are generated within an RTT, the smoothed RTT and RTT variance could retain inadequate history, as suggested in . Changing these computations is currently an open research question. min_rtt is the minimum RTT measured over the connection, prior to adjusting by ack delay. Ignoring ack delay for min RTT prevents intentional or unintentional underestimation of min RTT, which in turn prevents underestimating smoothed RTT.

QUIC senders use both ack information and timeouts to detect lost packets, and this section provides a description of these algorithms. If a packet is lost, the QUIC transport needs to recover from that loss, such as by retransmitting the data, sending an updated frame, or abandoning the frame. For more information, see Section 13.2 of .

Acknowledgement-based loss detection implements the spirit of TCP’s Fast Retransmit , Early Retransmit , FACK , SACK loss recovery , and RACK . This section provides an overview of how these algorithms are implemented in QUIC. A packet is declared lost if it meets all the following conditions: The packet is unacknowledged, in-flight, and was sent prior to an acknowledged packet. Either its packet number is kPacketThreshold smaller than an acknowledged packet (), or it was sent long enough in the past (). The acknowledgement indicates that a packet sent later was delivered, while the packet and time thresholds provide some tolerance for packet reordering. Spuriously declaring packets as lost leads to unnecessary retransmissions and may result in degraded performance due to the actions of the congestion controller upon detecting loss. Implementations that detect spurious retransmissions and increase the reordering threshold in packets or time MAY choose to start with smaller initial reordering thresholds to minimize recovery latency.

The RECOMMENDED initial value for the packet reordering threshold (kPacketThreshold) is 3, based on best practices for TCP loss detection . Some networks may exhibit higher degrees of reordering, causing a sender to detect spurious losses. Implementers MAY use algorithms developed for TCP, such as TCP-NCR , to improve QUIC’s reordering resilience.

Once a later packet has been acknowledged, an endpoint SHOULD declare an earlier packet lost if it was sent a threshold amount of time in the past. The time threshold is computed as kTimeThreshold * max(SRTT, latest_RTT). If packets sent prior to the largest acknowledged packet cannot yet be declared lost, then a timer SHOULD be set for the remaining time. The RECOMMENDED time threshold (kTimeThreshold), expressed as a round-trip time multiplier, is 9/8. Using max(SRTT, latest_RTT) protects from the two following cases: the latest RTT sample is lower than the SRTT, perhaps due to reordering where the acknowledgement encountered a shorter path; the latest RTT sample is higher than the SRTT, perhaps due to a sustained increase in the actual RTT, but the smoothed SRTT has not yet caught up. Implementations MAY experiment with absolute thresholds, thresholds from previous connections, adaptive thresholds, or including RTT variance. Smaller thresholds reduce reordering resilience and increase spurious retransmissions, and larger thresholds increase loss detection delay.

Data in CRYPTO frames is critical to QUIC transport and crypto negotiation, so a more aggressive timeout is used to retransmit it. The initial crypto retransmission timeout SHOULD be set to twice the initial RTT. At the beginning, there are no prior RTT samples within a connection. Resumed connections over the same network SHOULD use the previous connection’s final smoothed RTT value as the resumed connection’s initial RTT. If no previous RTT is available, or if the network changes, the initial RTT SHOULD be set to 100ms. When an acknowledgement is received, a new RTT is computed and the timer SHOULD be set for twice the newly computed smoothed RTT. When a crypto packet is sent, the sender MUST set a timer for the crypto timeout period. This timer MUST be updated when a new crypto packet is sent. Upon timeout, the sender MUST retransmit all unacknowledged CRYPTO data if possible. Until the server has validated the client’s address on the path, the amount of data it can send is limited, as specified in . If not all unacknowledged CRYPTO data can be sent, then all unacknowledged CRYPTO data sent in Initial packets should be retransmitted. If no data can be sent, then no alarm should be armed until data has been received from the client. Because the server could be blocked until more packets are received, the client MUST start the crypto retransmission timer even if there is no unacknowledged CRYPTO data. If the timer expires and the client has no CRYPTO data to retransmit and does not have Handshake keys, it SHOULD send an Initial packet in a UDP datagram of at least 1200 bytes. If the client has Handshake keys, it SHOULD send a Handshake packet. On each consecutive expiration of the crypto timer without receiving an acknowledgement for a new packet, the sender SHOULD double the crypto retransmission timeout and set a timer for this period. When crypto packets are in flight, the probe timer () is not active.

A Retry or Version Negotiation packet causes a client to send another Initial packet, effectively restarting the connection process and resetting congestion control and loss recovery state, including resetting any pending timers. Either packet indicates that the Initial was received but not processed. Neither packet can be treated as an acknowledgment for the Initial. The client MAY however compute an RTT estimate to the server as the time period from when the first Initial was sent to when a Retry or a Version Negotiation packet is received. The client MAY use this value to seed the RTT estimator for a subsequent connection attempt to the server.

When packet protection keys are discarded (see Section 4.9 of ), all packets that were sent with those keys can no longer be acknowledged because their acknowledgements cannot be processed anymore. The sender MUST discard all recovery state associated with those packets and MUST remove them from the count of bytes in flight. Endpoints stop sending and receiving Initial packets once they start exchanging Handshake packets (see Section 17.2.2.1 of ). At this point, recovery state for all in-flight Initial packets is discarded. When 0-RTT is rejected, recovery state for all in-flight 0-RTT packets is discarded. If a server accepts 0-RTT, but does not buffer 0-RTT packets that arrive before Initial packets, early 0-RTT packets will be declared lost, but that is expected to be infrequent. It is expected that keys are discarded after packets encrypted with them would be acknowledged or declared lost. Initial secrets however might be destroyed sooner, as soon as handshake keys are available (see Section 4.10 of ).

A Probe Timeout (PTO) triggers a probe packet when ack-eliciting data is in flight but an acknowledgement is not received within the expected period of time. A PTO enables a connection to recover from loss of tail packets or acks. The PTO algorithm used in QUIC implements the reliability functions of Tail Loss Probe , RTO and F-RTO algorithms for TCP , and the timeout computation is based on TCP’s retransmission timeout period .

When an ack-eliciting packet is transmitted, the sender schedules a timer for the PTO period as follows:

kGranularity, smoothed_rtt, rttvar, and max_ack_delay are defined in and . The PTO period is the amount of time that a sender ought to wait for an acknowledgement of a sent packet. This time period includes the estimated network roundtrip-time (smoothed_rtt), the variance in the estimate (4*rttvar), and max_ack_delay, to account for the maximum time by which a receiver might delay sending an acknowledgement. The PTO value MUST be set to at least kGranularity, to avoid the timer expiring immediately. When a PTO timer expires, the sender probes the network as described in the next section. The PTO period MUST be set to twice its current value. This exponential reduction in the sender’s rate is important because the PTOs might be caused by loss of packets or acknowledgements due to severe congestion. A sender computes its PTO timer every time an ack-eliciting packet is sent. A sender might choose to optimize this by setting the timer fewer times if it knows that more ack-eliciting packets will be sent within a short period of time.

When a PTO timer expires, the sender MUST send one ack-eliciting packet as a probe. A sender MAY send up to two ack-eliciting packets, to avoid an expensive consecutive PTO expiration due to a single packet loss. Consecutive PTO periods increase exponentially, and as a result, connection recovery latency increases exponentially as packets continue to be dropped in the network. Sending two packets on PTO expiration increases resilience to packet drops, thus reducing the probability of consecutive PTO events. Probe packets sent on a PTO MUST be ack-eliciting. A probe packet SHOULD carry new data when possible. A probe packet MAY carry retransmitted unacknowledged data when new data is unavailable, when flow control does not permit new data to be sent, or to opportunistically reduce loss recovery delay. Implementations MAY use alternate strategies for determining the content of probe packets, including sending new or retransmitted data based on the application’s priorities. When the PTO timer expires multiple times and new data cannot be sent, implementations must choose between sending the same payload every time or sending different payloads. Sending the same payload may be simpler and ensures the highest priority frames arrive first. Sending different payloads each time reduces the chances of spurious retransmission. When a PTO timer expires, new or previously-sent data may not be available to send and packets may still be in flight. A sender can be blocked from sending new data in the future if packets are left in flight. Under these conditions, a sender SHOULD mark any packets still in flight as lost. If a sender wishes to establish delivery of packets still in flight, it MAY send an ack-eliciting packet and re-arm the PTO timer instead.

Delivery or loss of packets in flight is established when an ACK frame is received that newly acknowledges one or more packets. A PTO timer expiration event does not indicate packet loss and MUST NOT cause prior unacknowledged packets to be marked as lost. When an acknowledgement is received that newly acknowledges packets, loss detection proceeds as dictated by packet and time threshold mechanisms, see .

The majority of constants were derived from best common practices among widely deployed TCP implementations on the internet. Exceptions follow. A shorter delayed ack time of 25ms was chosen because longer delayed acks can delay loss recovery and for the small number of connections where less than packet per 25ms is delivered, acking every packet is beneficial to congestion control and loss recovery. The default initial RTT of 100ms was chosen because it is slightly higher than both the median and mean min_rtt typically observed on the public internet.

QUIC’s congestion control is based on TCP NewReno . NewReno is a congestion window based congestion control. QUIC specifies the congestion window in bytes rather than packets due to finer control and the ease of appropriate byte counting . QUIC hosts MUST NOT send packets if they would increase bytes_in_flight (defined in ) beyond the available congestion window, unless the packet is a probe packet sent after a PTO timer expires, as described in . Implementations MAY use other congestion control algorithms, such as Cubic , and endpoints MAY use different algorithms from one another. The signals QUIC provides for congestion control are generic and are designed to support different algorithms.

If a path has been verified to support ECN, QUIC treats a Congestion Experienced codepoint in the IP header as a signal of congestion. This document specifies an endpoint’s response when its peer receives packets with the Congestion Experienced codepoint. As discussed in , endpoints are permitted to experiment with other response functions.

QUIC begins every connection in slow start and exits slow start upon loss or upon increase in the ECN-CE counter. QUIC re-enters slow start anytime the congestion window is less than ssthresh, which typically only occurs after an PTO. While in slow start, QUIC increases the congestion window by the number of bytes acknowledged when each acknowledgment is processed.

Slow start exits to congestion avoidance. Congestion avoidance in NewReno uses an additive increase multiplicative decrease (AIMD) approach that increases the congestion window by one maximum packet size per congestion window acknowledged. When a loss is detected, NewReno halves the congestion window and sets the slow start threshold to the new congestion window.

Recovery is a period of time beginning with detection of a lost packet or an increase in the ECN-CE counter. Because QUIC does not retransmit packets, it defines the end of recovery as a packet sent after the start of recovery being acknowledged. This is slightly different from TCP’s definition of recovery, which ends when the lost packet that started recovery is acknowledged. The recovery period limits congestion window reduction to once per round trip. During recovery, the congestion window remains unchanged irrespective of new losses or increases in the ECN-CE counter.

During the handshake, some packet protection keys might not be available when a packet arrives. In particular, Handshake and 0-RTT packets cannot be processed until the Initial packets arrive, and 1-RTT packets cannot be processed until the handshake completes. Endpoints MAY ignore the loss of Handshake, 0-RTT, and 1-RTT packets that might arrive before the peer has packet protection keys to process those packets.

Probe packets MUST NOT be blocked by the congestion controller. A sender MUST however count these packets as being additionally in flight, since these packets add network load without establishing packet loss. Note that sending probe packets might cause the sender’s bytes in flight to exceed the congestion window until an acknowledgement is received that establishes loss or delivery of packets.

When an ACK frame is received that establishes loss of all in-flight packets sent over a long enough period of time, the network is considered to be experiencing persistent congestion. Commonly, this can be established by consecutive PTOs, but since the PTO timer is reset when a new ack-eliciting packet is sent, an explicit duration must be used to account for those cases where PTOs do not occur or are substantially delayed. This duration is the equivalent of kPersistentCongestionThreshold consecutive PTOs, and is computed as follows: ~~~ (smoothed_rtt + 4 * rttvar + max_ack_delay) * ((2 ^ kPersistentCongestionThreshold) - 1) ~~~ For example, assume: smoothed_rtt = 1 rttvar = 0 max_ack_delay = 0 kPersistentCongestionThreshold = 2 If an eck-eliciting packet is sent at time = 0, the following scenario would illustrate persistent congestion: t=0 Send Pkt #1 (App Data) t=1 Send Pkt #2 (PTO 1) t=3 Send Pkt #3 (PTO 2) t=7 Send Pkt #4 (PTO 3) t=8 Recv ACK of Pkt #4 The first three packets are determined to be lost when the ACK of packet 4 is received at t=8. The congestion period is calculated as the time between the oldest and newest lost packets: (3 - 0) = 3. The duration for persistent congestion is equal to: (1 * ((2 ^ kPersistentCongestionThreshold) - 1)) = 3. Because the threshold was reached and because none of the packets between the oldest and the newest packets are acknowledged, the network is considered to have experienced persistent congestion. When persistent congestion is established, the sender’s congestion window MUST be reduced to the minimum congestion window (kMinimumWindow). This response of collapsing the congestion window on persistent congestion is functionally similar to a sender’s response on a Retransmission Timeout (RTO) in TCP after Tail Loss Probes (TLP) .

This document does not specify a pacer, but it is RECOMMENDED that a sender pace sending of all in-flight packets based on input from the congestion controller. For example, a pacer might distribute the congestion window over the SRTT when used with a window-based controller, and a pacer might use the rate estimate of a rate-based controller. An implementation should take care to architect its congestion controller to work well with a pacer. For instance, a pacer might wrap the congestion controller and control the availability of the congestion window, or a pacer might pace out packets handed to it by the congestion controller. Timely delivery of ACK frames is important for efficient loss recovery. Packets containing only ACK frames should therefore not be paced, to avoid delaying their delivery to the peer. As an example of a well-known and publicly available implementation of a flow pacer, implementers are referred to the Fair Queue packet scheduler (fq qdisc) in Linux (3.11 onwards).

A sender becomes idle if it ceases to send data and has no bytes in flight. A sender’s congestion window MUST NOT increase while it is idle. When sending data after becoming idle, a sender MUST reset its congestion window to the initial congestion window (see Section 4.1 of ), unless it paces the sending of packets. A sender MAY retain its congestion window if it paces the sending of any packets in excess of the initial congestion window. A sender MAY implement alternate mechanisms to update its congestion window after idle periods, such as those proposed for TCP in .

The congestion window should not be increased in slow start or congestion avoidance when it is not fully utilized. The congestion window could be under-utilized due to insufficient application data or flow control credit. A sender that paces packets (see ) might delay sending packets and not fully utilize the congestion window due to this delay. A sender should not consider itself application limited if it would have fully utilized the congestion window without pacing delay.

Congestion control fundamentally involves the consumption of signals – both loss and ECN codepoints – from unauthenticated entities. On-path attackers can spoof or alter these signals. An attacker can cause endpoints to reduce their sending rate by dropping packets, or alter send rate by changing ECN codepoints.

Packets that carry only ACK frames can be heuristically identified by observing packet size. Acknowledgement patterns may expose information about link characteristics or application behavior. Endpoints can use PADDING frames or bundle acknowledgments with other frames to reduce leaked information.

A receiver can misreport ECN markings to alter the congestion response of a sender. Suppressing reports of ECN-CE markings could cause a sender to increase their send rate. This increase could result in congestion and loss. A sender MAY attempt to detect suppression of reports by marking occasional packets that they send with ECN-CE. If a packet marked with ECN-CE is not reported as having been marked when the packet is acknowledged, the sender SHOULD then disable ECN for that path. Reporting additional ECN-CE markings will cause a sender to reduce their sending rate, which is similar in effect to advertising reduced connection flow control limits and so no advantage is gained by doing so. Endpoints choose the congestion controller that they use. Though congestion controllers generally treat reports of ECN-CE markings as equivalent to loss , the exact response for each controller could be different. Failure to correctly respond to information about ECN markings is therefore difficult to detect.

This document has no IANA actions. Yet.

Fastly Mozilla Mozilla sn3rd In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings. This memo updates RFC 3168, which specifies Explicit Congestion Notification (ECN) as an alternative to packet drops for indicating network congestion to endpoints. It relaxes restrictions in RFC 3168 that hinder experimentation towards benefits beyond just removal of loss. This memo summarizes the anticipated areas of experimentation and updates RFC 3168 to enable experimentation in these areas. An Experimental RFC in the IETF document stream is required to take advantage of any of these enabling updates. In addition, this memo makes related updates to the ECN specifications for RTP in RFC 6679 and for the Datagram Congestion Control Protocol (DCCP) in RFCs 4341, 4342, and 5622. This memo also records the conclusion of the ECN nonce experiment in RFC 3540 and provides the rationale for reclassification of RFC 3540 from Experimental to Historic; this reclassification enables new experimental use of the ECT(1) codepoint. This document defines the standard algorithm that Transmission Control Protocol (TCP) senders are required to use to compute and manage their retransmission timer. It expands on the discussion in Section 4.2.3.1 of RFC 1122 and upgrades the requirement of supporting the algorithm from a SHOULD to a MUST. This document obsoletes RFC 2988. [STANDARDS-TRACK] This document defines TCP's four intertwined congestion control algorithms: slow start, congestion avoidance, fast retransmit, and fast recovery. In addition, the document specifies how TCP should begin transmission after a relatively long idle period, as well as discussing various acknowledgment generation methods. This document obsoletes RFC 2581. [STANDARDS-TRACK] This document proposes a new mechanism for TCP and Stream Control Transmission Protocol (SCTP) that can be used to recover lost segments when a connection's congestion window is small. The "Early Retransmit" mechanism allows the transport to reduce, in certain special circumstances, the number of duplicate acknowledgments required to trigger a fast retransmission. This allows the transport to use fast retransmit to recover segment losses that would otherwise require a lengthy retransmission timeout. [STANDARDS-TRACK] This document presents a conservative loss recovery algorithm for TCP that is based on the use of the selective acknowledgment (SACK) TCP option. The algorithm presented in this document conforms to the spirit of the current congestion control specification (RFC 5681), but allows TCP senders to recover more effectively when multiple segments are lost from a single flight of data. This document obsoletes RFC 3517 and describes changes from it. [STANDARDS-TRACK] This document presents a new TCP loss detection algorithm called RACK ("Recent ACKnowledgment"). RACK uses the notion of time, instead of packet or sequence counts, to detect losses, for modern TCP implementations that can support per-packet timestamps and the selective acknowledgment (SACK) option. It is intended to replace the conventional DUPACK threshold approach and its variants, as well as other nonstandard approaches. This document specifies Non-Congestion Robustness (NCR) for TCP. In the absence of explicit congestion notification from the network, TCP uses loss as an indication of congestion. One of the ways TCP detects loss is using the arrival of three duplicate acknowledgments. However, this heuristic is not always correct, notably in the case when network paths reorder segments (for whatever reason), resulting in degraded performance. TCP-NCR is designed to mitigate this degraded performance by increasing the number of duplicate acknowledgments required to trigger loss recovery, based on the current state of the connection, in an effort to better disambiguate true segment loss from segment reordering. This document specifies the changes to TCP, as well as the costs and benefits of these modifications. This memo defines an Experimental Protocol for the Internet community. Retransmission timeouts are detrimental to application latency, especially for short transfers such as Web transactions where timeouts can often take longer than all of the rest of a transaction. The primary cause of retransmission timeouts are lost segments at the tail of transactions. This document describes an experimental algorithm for TCP to quickly recover lost segments at the end of transactions or when an entire window of data or acknowledgments are lost. Tail Loss Probe (TLP) is a sender-only algorithm that allows the transport to recover tail losses through fast recovery as opposed to lengthy retransmission timeouts. If a connection is not receiving any acknowledgments for a certain period of time, TLP transmits the last unacknowledged segment (loss probe). In the event of a tail loss in the original transmissions, the acknowledgment from the loss probe triggers SACK/FACK based fast recovery. TLP effectively avoids long timeouts and thereby improves TCP performance. The purpose of this document is to move the F-RTO (Forward RTO-Recovery) functionality for TCP in RFC 4138 from Experimental to Standards Track status. The F-RTO support for Stream Control Transmission Protocol (SCTP) in RFC 4138 remains with Experimental status. See Appendix B for the differences between this document and RFC 4138.Spurious retransmission timeouts cause suboptimal TCP performance because they often result in unnecessary retransmission of the last window of data. This document describes the F-RTO detection algorithm for detecting spurious TCP retransmission timeouts. F-RTO is a TCP sender-only algorithm that does not require any TCP options to operate. After retransmitting the first unacknowledged segment triggered by a timeout, the F-RTO algorithm of the TCP sender monitors the incoming acknowledgments to determine whether the timeout was spurious. It then decides whether to send new segments or retransmit unacknowledged segments. The algorithm effectively helps to avoid additional unnecessary retransmissions and thereby improves TCP performance in the case of a spurious timeout. [STANDARDS-TRACK] RFC 5681 documents the following four intertwined TCP congestion control algorithms: slow start, congestion avoidance, fast retransmit, and fast recovery. RFC 5681 explicitly allows certain modifications of these algorithms, including modifications that use the TCP Selective Acknowledgment (SACK) option (RFC 2883), and modifications that respond to "partial acknowledgments" (ACKs that cover new data, but not all the data outstanding when loss was detected) in the absence of SACK. This document describes a specific algorithm for responding to partial acknowledgments, referred to as "NewReno". This response to partial acknowledgments was first proposed by Janey Hoe. This document obsoletes RFC 3782. [STANDARDS-TRACK] This document proposes a small modification to the way TCP increases its congestion window. Rather than the traditional method of increasing the congestion window by a constant amount for each arriving acknowledgment, the document suggests basing the increase on the number of previously unacknowledged bytes each ACK covers. This change improves the performance of TCP, as well as closes a security hole TCP receivers can use to induce the sender into increasing the sending rate too rapidly. This memo defines an Experimental Protocol for the Internet community. CUBIC is an extension to the current TCP standards. It differs from the current TCP standards only in the congestion control algorithm on the sender side. In particular, it uses a cubic function instead of a linear window increase function of the current TCP standards to improve scalability and stability under fast and long-distance networks. CUBIC and its predecessor algorithm have been adopted as defaults by Linux and have been used for many years. This document provides a specification of CUBIC to enable third-party implementations and to solicit community feedback through experimentation on the performance of CUBIC. This document provides a mechanism to address issues that arise when TCP is used for traffic that exhibits periods where the sending rate is limited by the application rather than the congestion window. It provides an experimental update to TCP that allows a TCP sender to restart quickly following a rate-limited interval. This method is expected to benefit applications that send rate-limited traffic using TCP while also providing an appropriate response if congestion is experienced.This document also evaluates the Experimental specification of TCP Congestion Window Validation (CWV) defined in RFC 2861 and concludes that RFC 2861 sought to address important issues but failed to deliver a widely used solution. This document therefore reclassifies the status of RFC 2861 from Experimental to Historic. This document obsoletes RFC 2861. This document proposes an experiment to increase the permitted TCP initial window (IW) from between 2 and 4 segments, as specified in RFC 3390, to 10 segments with a fallback to the existing recommendation when performance issues are detected. It discusses the motivation behind the increase, the advantages and disadvantages of the higher initial window, and presents results from several large-scale experiments showing that the higher initial window improves the overall performance of many web services without resulting in a congestion collapse. The document closes with a discussion of usage and deployment for further experimental purposes recommended by the IETF TCP Maintenance and Minor Extensions (TCPM) working group.

We now describe an example implementation of the loss detection mechanisms described in .

To correctly implement congestion control, a QUIC sender tracks every ack-eliciting packet until the packet is acknowledged or lost. It is expected that implementations will be able to access this information by packet number and crypto context and store the per-packet fields () for loss recovery and congestion control. After a packet is declared lost, it SHOULD be tracked for an amount of time comparable to the maximum expected packet reordering, such as 1 RTT. This allows for detection of spurious retransmissions. Sent packets are tracked for each packet number space, and ACK processing only applies to a single space.

The packet number of the sent packet. A boolean that indicates whether a packet is ack-eliciting. If true, it is expected that an acknowledgement will be received, though the peer could delay sending the ACK frame containing it by up to the MaxAckDelay. A boolean that indicates whether the packet counts towards bytes in flight. A boolean that indicates whether the packet contains cryptographic handshake messages critical to the completion of the QUIC handshake. In this version of QUIC, this includes any packet with the long header that includes a CRYPTO frame. The number of bytes sent in the packet, not including UDP or IP overhead, but including QUIC framing overhead. The time the packet was sent.

Constants used in loss recovery are based on a combination of RFCs, papers, and common practice. Some may need to be changed or negotiated in order to better suit a variety of environments. Maximum reordering in packets before packet threshold loss detection considers a packet lost. The RECOMMENDED value is 3. Maximum reordering in time before time threshold loss detection considers a packet lost. Specified as an RTT multiplier. The RECOMMENDED value is 9/8. Timer granularity. This is a system-dependent value. However, implementations SHOULD use a value no smaller than 1ms. The RTT used before an RTT sample is taken. The RECOMMENDED value is 100ms. An enum to enumerate the three packet number spaces. ~~~ enum kPacketNumberSpace { Initial, Handshake, ApplicationData, } ~~~

Variables required to implement the congestion control mechanisms are described in this section. Multi-modal timer used for loss detection. The number of times all unacknowledged CRYPTO data has been retransmitted without receiving an ack. The number of times a PTO has been sent without receiving an ack. The time the most recent ack-eliciting packet was sent. The time the most recent crypto packet was sent. The largest packet number acknowledged in the packet number space so far. The most recent RTT measurement made when receiving an ack for a previously unacked packet. The smoothed RTT of the connection, computed as described in The RTT variance, computed as described in The minimum RTT seen in the connection, ignoring ack delay. The maximum amount of time by which the receiver intends to delay acknowledgments, in milliseconds. The actual ack_delay in a received ACK frame may be larger due to late timers, reordering, or lost ACKs. The time at which the next packet in that packet number space will be considered lost based on exceeding the reordering window in time. An association of packet numbers in a packet number space to information about them. Described in detail above in .

At the beginning of the connection, initialize the loss detection variables as follows:

After a packet is sent, information about the packet is stored. The parameters to OnPacketSent are described in detail above in . Pseudocode for OnPacketSent follows:

When an ACK frame is received, it may newly acknowledge any number of packets. Pseudocode for OnAckReceived and UpdateRtt follow:

ack_delay): latest_rtt -= ack_delay // Based on {{RFC6298}}. if (smoothed_rtt == 0): smoothed_rtt = latest_rtt rttvar = latest_rtt / 2 else: rttvar_sample = abs(smoothed_rtt - latest_rtt) rttvar = 3/4 * rttvar + 1/4 * rttvar_sample smoothed_rtt = 7/8 * smoothed_rtt + 1/8 * latest_rtt ]]>

When a packet is acknowledged for the first time, the following OnPacketAcked function is called. Note that a single ACK frame may newly acknowledge several packets. OnPacketAcked must be called once for each of these newly acknowledged packets. OnPacketAcked takes two parameters: acked_packet, which is the struct detailed in , and the packet number space that this ACK frame was sent for. Pseudocode for OnPacketAcked follows:

QUIC loss detection uses a single timer for all timeout loss detection. The duration of the timer is based on the timer’s mode, which is set in the packet and timer events further below. The function SetLossDetectionTimer defined below shows how the single timer is set. This algorithm may result in the timer being set in the past, particularly if timers wake up late. Timers set in the past SHOULD fire immediately. Pseudocode for SetLossDetectionTimer follows:

When the loss detection timer expires, the timer’s mode determines the action to be performed. Pseudocode for OnLossDetectionTimeout follows:

DetectLostPackets is called every time an ACK is received and operates on the sent_packets for that packet number space. If the loss detection timer expires and the loss_time is set, the previous largest acknowledged packet is supplied. Pseudocode for DetectLostPackets follows:

largest_acked_packet[pn_space]): continue // Mark packet as lost, or set time when it should be marked. if (unacked.time_sent <= lost_send_time || unacked.packet_number <= lost_pn): sent_packets.remove(unacked.packet_number) if (unacked.in_flight): lost_packets.insert(unacked) else: if (loss_time[pn_space] == 0): loss_time[pn_space] = unacked.time_sent + loss_delay else: loss_time[pn_space] = min(loss_time[pn_space], unacked.time_sent + loss_delay) // Inform the congestion controller of lost packets and // let it decide whether to retransmit immediately. if (!lost_packets.empty()): OnPacketsLost(lost_packets) ]]>

We now describe an example implementation of the congestion controller described in .

Constants used in congestion control are based on a combination of RFCs, papers, and common practice. Some may need to be changed or negotiated in order to better suit a variety of environments. The sender’s maximum payload size. Does not include UDP or IP overhead. The max packet size is used for calculating initial and minimum congestion windows. The RECOMMENDED value is 1200 bytes. Default limit on the initial amount of data in flight, in bytes. Taken from , but increased slightly to account for the smaller 8 byte overhead of UDP vs 20 bytes for TCP. The RECOMMENDED value is the minimum of 10 * kMaxDatagramSize and max(2* kMaxDatagramSize, 14720)). Minimum congestion window in bytes. The RECOMMENDED value is 2 * kMaxDatagramSize. Reduction in congestion window when a new loss event is detected. The RECOMMENDED value is 0.5. Number of consecutive PTOs required for persistent congestion to be established. The rationale for this threshold is to enable a sender to use initial PTOs for aggressive probing, as TCP does with Tail Loss Probe (TLP) , before establishing persistent congestion, as TCP does with a Retransmission Timeout (RTO) . The RECOMMENDED value for kPersistentCongestionThreshold is 2, which is equivalent to having two TLPs before an RTO in TCP.

Variables required to implement the congestion control mechanisms are described in this section. The highest value reported for the ECN-CE counter by the peer in an ACK frame. This variable is used to detect increases in the reported ECN-CE counter. The sum of the size in bytes of all sent packets that contain at least one ack-eliciting or PADDING frame, and have not been acked or declared lost. The size does not include IP or UDP overhead, but does include the QUIC header and AEAD overhead. Packets only containing ACK frames do not count towards bytes_in_flight to ensure congestion control does not impede congestion feedback. Maximum number of bytes-in-flight that may be sent. The time when QUIC first detects a loss, causing it to enter recovery. When a packet sent after this time is acknowledged, QUIC exits recovery. Slow start threshold in bytes. When the congestion window is below ssthresh, the mode is slow start and the window grows by the number of bytes acknowledged.

At the beginning of the connection, initialize the congestion control variables as follows:

Whenever a packet is sent, and it contains non-ACK frames, the packet increases bytes_in_flight.

Invoked from loss detection’s OnPacketAcked and is supplied with the acked_packet from sent_packets.

Invoked from ProcessECN and OnPacketsLost when a new congestion event is detected. May start a new recovery period and reduces the congestion window.

Invoked when an ACK frame with an ECN section is received from the peer.

ecn_ce_counter): ecn_ce_counter = ack.ce_counter // Start a new congestion event if the last acknowledged // packet was sent after the start of the previous // recovery epoch. CongestionEvent(sent_packets[ack.largest_acked].time_sent) ]]>

Invoked by loss detection from DetectLostPackets when new packets are detected lost.

RFC Editor’s Note: Please remove this section prior to publication of a final version of this document. Issue and pull request numbers are listed with a leading octothorp.

Change IW byte limit to 14720 from 14600 (#2494) Update PTO calculation to match RFC6298 (#2480, #2489, #2490) Improve loss detection’s description of multiple packet number spaces and pseudocode (#2485, #2451, #2417) Declare persistent congestion even if non-probe packets are sent and don’t make persistent congestion more aggressive than RTO verified was (#2365, #2244) Move pseudocode to the appendices (#2408) What to send on multiple PTOs (#2380)

After Probe Timeout discard in-flight packets or send another (#2212, #1965) Endpoints discard initial keys as soon as handshake keys are available (#1951, #2045) 0-RTT state is discarded when 0-RTT is rejected (#2300) Loss detection timer is cancelled when ack-eliciting frames are in flight (#2117, #2093) Packets are declared lost if they are in flight (#2104) After becoming idle, either pace packets or reset the congestion controller (#2138, 2187) Process ECN counts before marking packets lost (#2142) Mark packets lost before resetting crypto_count and pto_count (#2208, #2209) Congestion and loss recovery state are discarded when keys are discarded (#2327)

Unify TLP and RTO into a single PTO; eliminate min RTO, min TLP and min crypto timeouts; eliminate timeout validation (#2114, #2166, #2168, #1017) Redefine how congestion avoidance in terms of when the period starts (#1928, #1930) Document what needs to be tracked for packets that are in flight (#765, #1724, #1939) Integrate both time and packet thresholds into loss detection (#1969, #1212, #934, #1974) Reduce congestion window after idle, unless pacing is used (#2007, #2023) Disable RTT calculation for packets that don’t elicit acknowledgment (#2060, #2078) Limit ack_delay by max_ack_delay (#2060, #2099) Initial keys are discarded once Handshake are avaialble (#1951, #2045) Reorder ECN and loss detection in pseudocode (#2142) Only cancel loss detection timer if ack-eliciting packets are in flight (#2093, #2117)

Used max_ack_delay from transport params (#1796, #1782) Merge ACK and ACK_ECN (#1783)

Corrected the lack of ssthresh reduction in CongestionEvent pseudocode (#1598) Considerations for ECN spoofing (#1426, #1626) Clarifications for PADDING and congestion control (#837, #838, #1517, #1531, #1540) Reduce early retransmission timer to RTT/8 (#945, #1581) Packets are declared lost after an RTO is verified (#935, #1582)

Changes to manage separate packet number spaces and encryption levels (#1190, #1242, #1413, #1450) Added ECN feedback mechanisms and handling; new ACK_ECN frame (#804, #805, #1372)

No significant changes.

Improved text on ack generation (#1139, #1159) Make references to TCP recovery mechanisms informational (#1195) Define time_of_last_sent_handshake_packet (#1171) Added signal from TLS the data it includes needs to be sent in a Retry packet (#1061, #1199) Minimum RTT (min_rtt) is initialized with an infinite value (#1169)

No significant changes.

Clarified pacing and RTO (#967, #977)

Include Ack Delay in RTO(and TLP) computations (#981) Ack Delay in SRTT computation (#961) Default RTT and Slow Start (#590) Many editorial fixes.

No significant changes.

Add more congestion control text (#776)

No significant changes.

No significant changes.

Integrate F-RTO (#544, #409) Add congestion control (#545, #395) Require connection abort if a skipped packet was acknowledged (#415) Simplify RTO calculations (#142, #417)

Overview added to loss detection Changes initial default RTT to 100ms Added time-based loss detection and fixes early retransmit Clarified loss recovery for handshake packets Fixed references and made TCP references informative

Improved description of constants and ACK behavior

Adopted as base for draft-ietf-quic-recovery Updated authors/editors list Added table of contents