Device, method and system to enforce concurrency limits of a target node within a network fabric

Patent metadata

Field	Value
Publication number	US-2018287904-A1
Application number	US-201715474833-A
Country	US
Kind code	A1
Filing date	Mar 30, 2017
Priority date	Mar 30, 2017
Publication date	Oct 4, 2018
Grant date	—

Abstract

A computing device, a method, and a system to enforce concurrency limits within a network fabric. The computing device includes a memory device; and a network interface controller coupled to the memory device. The network interface controller includes circuitry to communicate with a plurality of target computing devices in a network fabric. The circuitry is configured to generate packets for transmission to respective ones of a plurality of target nodes. For each packet addressed to its corresponding target node, the circuitry is to determine whether transmitting the packet would violate the target node's concurrency limit. If transmitting the packet would not violate the target node's concurrency limit, the circuitry is to transmit the packet to the target node. However, if transmitting the packet would violate the target node's concurrency limit, the circuitry would store the packet in a retransmit buffer of the computing device, where the retransmit buffer is to further store already transmitted packets flagged for retransmission. The circuitry would then transmit the packet from the retransmit buffer when transmitting the packet from the retransmit buffer would not violate the target node's concurrency limit.

First claim

What is claimed is:

1 . A computing device including:
a memory device; and a network interface controller coupled to the memory device and including circuitry to communicate with a plurality of target nodes in a network fabric, the circuitry to: generate packets for transmission to respective ones of the target nodes; for each packet addressed to its corresponding target node:
determine whether transmitting the packet would violate the target node's concurrency limit;
if transmitting the packet would not violate the target node's concurrency limit, transmit the packet to the target node;
if transmitting the packet would violate the target node's concurrency limit:
store the packet in a retransmit buffer of the computing device; and
transmit the packet from the retransmit buffer when transmitting the packet from the retransmit buffer would not violate the target node's concurrency limit.

2 . The computing device of claim 1 , wherein the circuitry to generate, for transmission to each of the plurality of target nodes, a stream of contiguously sequenced packets in a packet sequence number (PSN) space.

3 . The computing device of claim 2 , wherein the concurrency limit corresponds to a Maximum Sequence Distance Limit (MSDL) of the target node.

4 . The computing device of claim 1 , wherein the concurrency limit corresponds to a Maximum Number of Missing Packets (MNMP) for the target node.

5 . The computing device of claim 1 , wherein the circuitry is to receive from the target node a first indication regarding received packets, and a second indication regarding packets not received, the circuitry to determine whether transmitting the packet would violate the target node's concurrency limit based on the first and second indications.

6 . The computing device of claim 5 , wherein the circuitry is to receive from the target node a third indication regarding the target node's concurrency limit, the circuitry to determine whether transmitting the packet would violate the target node's concurrency limit based on the first, second and third indications.

7 . The computing device of claim 1 , wherein the circuitry is to flag the packet as an out-of-range packet as part of storing the packet in the retransmit buffer.

8 . The computing device of claim 1 , wherein the computing device is to move the packet stored in the retransmit buffer to an in-window state before transmitting the packet.

9 . A method to transmit packets from a computing device in a network fabric, the method including:
generating packets for transmission to respective ones of a plurality of target nodes; for each packet addressed to its corresponding target node:
determining whether transmitting the packet would violate the target node's concurrency limit;
if transmitting the packet would not violate the target node's concurrency limit, transmitting the packet to the target node;
if transmitting the packet would violate the target node's concurrency limit:
storing the packet in a retransmit buffer of the computing device, the retransmit buffer to further store already transmitted packets flagged for retransmission;
transmitting the packet from the retransmit buffer when transmitting the packet from the retransmit buffer would not violate the target node's concurrency limit.

10 . The method of claim 9 , further including generating, for transmission to each of the plurality of target nodes, a stream of contiguously sequenced packets in a packet sequence number (PSN) space.

11 . The method of claim 10 , wherein the concurrency limit corresponds to a Maximum Sequence Distance Limit (MSDL) of the target node.

12 . The method of claim 9 , wherein the concurrency limit corresponds to a Maximum Number of Missing Packets (MNMP) for the target node.

13 . The method of claim 9 , further including:
receiving from the target node a first indication regarding received packets, and a second indication regarding packets not received; and determining whether transmitting the packet would violate the target node's concurrency limit based on the first and second indications.

14 . The method of claim 13 , further including:
receiving from the target node a third indication regarding the target node's concurrency limit; and determining whether transmitting the packet would violate the target node's concurrency limit based on the first, second and third indications.

15 . The method of claim 9 , wherein storing includes flagging the packet as an out-of-range packet.

16 . The method of claim 9 , further including moving the packet stored in the retransmit buffer to an in-window state before transmitting the packet.

17 . A product comprising one or more tangible computer-readable non-transitory storage media comprising computer-executable instructions operable to, when executed by at least one computer processor, enable the at least one computer processor to implement operations at a computing device, the operations including:
generating packets for transmission to respective ones of a plurality of target nodes; for each packet addressed to its corresponding target node:
determining whether transmitting the packet would violate the target node's concurrency limit;
if transmitting the packet would not violate the target node's concurrency limit, transmitting the packet to the target node;
if transmitting the packet would violate the target node's concurrency limit:
storing the packet in a retransmit buffer of the computing device, the retransmit buffer to further store already transmitted packets flagged for retransmission;
transmitting the packet from the retransmit buffer when transmitting the packet from the retransmit buffer would not violate the target node's concurrency limit.

18 . The product of claim 17 , wherein the operations further include generating, for transmission to each of the plurality of target nodes, a stream of contiguously sequenced packets in a packet sequence number (PSN) space, and wherein the concurrency limit corresponds to a Maximum Sequence Distance Limit (MSDL) of the target node.

19 . The product of claim 17 , wherein the concurrency limit corresponds to a Maximum Number of Missing Packets (MNMP) for the target node.

20 . The product of claim 17 , wherein the operations further include:
receiving from the target node a first indication regarding received packets, and a second indication regarding packets not received; and determining whether transmitting the packet would violate the target node's concurrency limit based on the first and second indications.

21 . The product of claim 20 , wherein the operations further include:
receiving from the target node a third indication regarding the target node's concurrency limit; and determining whether transmitting the packet would violate the target node's concurrency limit based on the first, second and third indications.

22 . The product of claim 17 , wherein the operations further include flagging the packet as an out-of-range packet.

23 . A system comprising:
a first computing device; and a plurality of second computing devices connected to the first computing device through a network fabric; wherein:
the first computing device comprises a network interface controller to:
generate packets for transmission to respective ones of the plurality of second computing devices;
for each packet addressed to its corresponding second computing device:
determine wh

Assignees

• Intel Corp

Inventors

• Brummel Karl P
• Giefer Charles A
• Miller Nathan S
• Underwood Keith D

Classifications

• H04L43/16

  Threshold monitoring · CPC title

• H04L67/10

  in which an application is distributed across nodes in the network (software deployment G06F8/60; multiprogramming arrangements G06F9/46) · CPC title

• H04L43/062 Primary

  related to network traffic · CPC title

• H04L47/10

  Flow control; Congestion control · CPC title

• H04L47/24 Primary

  Traffic characterised by specific attributes, e.g. priority or QoS · CPC title