Programmable congestion control communication scheme

Patent metadata

Field	Value
Publication number	US-11296988-B2
Application number	US-202016986428-A
Country	US
Kind code	B2
Filing date	Aug 6, 2020
Priority date	Nov 14, 2019
Publication date	Apr 5, 2022
Grant date	Apr 5, 2022

Abstract

A network adapter includes a receive (Rx) pipeline, a transmit (Tx) pipeline and congestion management circuitry. The Rx pipeline is configured to receive packets sent over a network by a peer network adapter, and to process the received packets. The Tx pipeline is configured to transmit packets to the peer network adapter over the network. The congestion management circuitry is configured to receive, from the Tx pipeline and from the Rx pipeline, Congestion-Control (CC) events derived from at least some of the packets exchanged with the peer network adapter, to exchange user-programmable congestion control packets with the peer network adapter, and to mitigate a congestion affecting one or more of the packets responsively to the CC events and the user-programmable congestion control packets.

First claim

The invention claimed is:

1. A network adapter, comprising:
a receive (Rx) pipeline, configured to receive packets sent over a network by a peer network adapter, and to process the received packets;
a transmit (Tx) pipeline, configured to transmit packets to the peer network adapter over the network; and
congestion management circuitry, configured to:
receive user-defined Congestion-Control (CC) software code that executes a CC algorithm selected from among multiple different CC algorithms, wherein the selected CC algorithm specifies user-programmable congestion control packets to be exchanged between the congestion management circuitry and peer congestion management circuitry in the peer network adapter;
receive, from the Tx pipeline and from the Rx pipeline, CC events derived from at least some of the packets exchanged with the peer network adapter;
exchange the user-programmable congestion control packets specified in the selected CC algorithm with the peer congestion management circuitry in the peer network adapter; and
mitigate a congestion affecting one or more of the packets responsively to the CC events and the user-programmable congestion control packets, by executing the selected CC algorithm.

2. The network adapter according to claim 1 , wherein the congestion affects the packets transmitted to the peer network adapter, and wherein the congestion management circuitry is configured to throttle, depending on the CC events and the user-programmable congestion control packets, a transmission rate of the packets transmitted by the Tx pipeline.

3. The network adapter according to claim 1 , wherein the congestion affects the packets received from the peer network adapter, and wherein the congestion management circuitry is configured to generate a congestion notification based on the CC events and the user-programmable congestion control packets, and to send the congestion notification to the peer network adapter.

4. The network adapter according to claim 1 , wherein the congestion management circuitry supports an Application Programming Interface (API) for receiving the user-defined Congestion-Control (CC) software code.

5. The network adapter according to claim 1 , wherein the programmable congestion control packets have a format of Management Datagram (MAD) packets.

6. The network adapter according to claim 1 , wherein the congestion management circuitry comprises a hardware congestion-control-circuitry configured to pre-process the CC events, and a user-programmable congestion control processor configured to mitigate the congestion responsively to the pre-processed CC events.

7. The network adapter according to claim 6 , wherein the hardware congestion-control-circuitry is configured to coalesce two or more of the CC events, and to provide the coalesced CC events to the congestion control processor.

8. The network adapter according to claim 1 , wherein the user-programmable congestion control packets have a payload format, which supports user-defined codes that coordinate execution of the selected CC algorithm between the congestion management circuitry and the congestion management circuitry in the peer network adapter.

9. A method for communication, comprising:
using a receive (Rx) pipeline of a network adapter, receiving packets sent over a network by a peer network adapter, and processing the received packets;
using a transmit (Tx) pipeline of the network adapter, transmitting packets to the peer network adapter over the network; and
using congestion management circuitry in the network adapter:
receiving user-defined Congestion-Control (CC) software code that executes a CC algorithm selected from among multiple different CC algorithms, wherein the selected CC algorithm specifies user-programmable congestion control packets to be exchanged between the congestion management circuitry and peer congestion management circuitry in the peer network adapter;
receiving from the Tx pipeline and from the Rx pipeline CC events derived from at least some of the packets exchanged with the peer network adapter;
exchanging the user-programmable congestion control packets specified in the selected CC algorithm with the peer congestion management circuitry in the peer network adapter; and
mitigating a congestion affecting one or more of the packets responsively to the CC events and the user-programmable congestion control packets, by executing the selected CC algorithm.

10. The method according to claim 9 , wherein the congestion affects the packets transmitted to the peer network adapter, and wherein mitigating the congestion comprises throttling, depending on the CC events and the user-programmable congestion control packets, a transmission rate of the packets transmitted by the Tx pipeline.

11. The method according to claim 9 , wherein the congestion affects the packets received from the peer network adapter, and wherein mitigating the congestion comprises generating a congestion notification based on the CC events and the user-programmable congestion control packets, and sending the congestion notification to the peer network adapter.

12. The method according to claim 9 , and comprising receiving the user-defined Congestion-Control (CC) software code in the congestion management circuitry using an Application Programming Interface (API).

13. The method according to claim 9 , wherein the programmable congestion control packets have a format of Management Datagram (MAD) packets.

14. The method according to claim 9 , wherein mitigating the congestion comprises pre-processing the CC events using hardware congestion-control-circuitry, and mitigating the congestion responsively to the pre-processed CC events by a user-programmable congestion control processor.

15. The method according to claim 14 , wherein pre-processing the CC events comprises coalescing two or more of the CC events, and providing the coalesced CC events to the congestion control processor.

16. The method according to claim 9 , wherein the user-programmable congestion control packets have a payload format, which supports user-defined codes that coordinate execution of the selected CC algorithm between the congestion management circuitry and the congestion management circuitry in the peer network adapter.

17. A network adapter, comprising:
a receive (Rx) pipeline, configured to receive packets sent over a network by a peer network adapter, and to process the received packets;
a transmit (Tx) pipeline, configured to transmit packets to the peer network adapter over the network; and
management circuitry, configured to:
receive user-defined software code that executes a control algorithm selected from among multiple different control algorithms, wherein the selected control algorithm specifies user-programmable control packets to be exchanged between the management circuitry and peer management circuitry in the peer network adapter;
receive, from the Tx pipeline and from the Rx pipeline, events derived from at least some of the packets exchanged with the peer network adapter;
exchange the user-programmable control packets specified in the selected control algorithm with the peer management circuitry in the peer network adapter, and
perform a control operation responsively to the events and the user-programmable control packets, by executing the selected control algorithm.

18. A method for communication, comprising:
using a receive (Rx) pipeline of a network adapter, receiving packets sent over a network by a peer network adapter, and processing the received packets;
using a transmi

Assignees

• Mellanox Technologies Ltd

Inventors

• Shpigelman Yuval
• Burstein Idan
• Bloch Noam
• Zuck Reut
• Moyal Roee

Classifications

• H04L47/12 Primary

  Avoiding congestion; Recovering from congestion · CPC title

• H04L47/267

  sent by the destination endpoint (network streaming of media packets with control of the source by the destination H04L65/613) · CPC title

• H04L47/26

  using explicit feedback to the source, e.g. choke packets · CPC title

• H04L47/25 Primary

  with rate being modified by the source upon detecting a change of network conditions · CPC title

• H04L69/22

  Parsing or analysis of headers · CPC title