System and method to provide multicast group membership defined relative to partition membership in a high performance computing environment

Patent metadata

Field	Value
Publication number	US-11949530-B2
Application number	US-202117531607-A
Country	US
Kind code	B2
Filing date	Nov 19, 2021
Priority date	Mar 24, 2017
Publication date	Apr 2, 2024
Grant date	Apr 2, 2024

Abstract

Systems and methods for providing multicast group (MCG) membership relative to partition membership in a high performance computing environment. In allowing a subnet manager of a local subnet to be instructed that all ports that are members of the relevant partition should be set up as members for a specific multicast group, the SM can perform a more efficient multicast-routing process. It is also possible to limit the IB client interaction with subnet administration conventionally required to handle join and leave operations. Additionally, subnet manager overhead can be reduced by creating a spanning tree for the routing of multicast packets that includes each of the partition members added to the multicast group, instead of creating a spanning tree after each multicast group join request is received, as conventionally required.

First claim

What is claimed is:

1. A method for providing multicast group membership defined relative to partition membership in a high performance computing environment, comprising:
providing, at a computer comprising one or more microprocessors, a subnet, the subnet comprising one or more switches, a plurality of end nodes interconnected via the one or more switches, and a subnet manager running on a node of the subnet;
generating, at an end node of the subnet, a multicast join request message, wherein the multicast join message comprises a plurality of subnet administration attributes, wherein the end node is a member of a partition defined within the subnet, wherein a set of the plurality of end nodes, including the end node, are defined as members of the partition; and
indicating, via one of the plurality of subnet administration attributes of the multicast join message, that each of the set of the plurality of end nodes should be added to a same multicast group.

2. The method of claim 1 , further comprising:
indicating, via another of the plurality of subnet administration attributes of the multicast join message, an identifier of the partition.

3. The method of claim 2 , further comprising:
receiving, at the subnet manager, the multicast join request message; and
upon receiving the multicast join request message, adding all members of the partition to a same multicast group.

4. The method of claim 3 , further comprising:
generating a spanning tree, the spanning tree including all end nodes of the subnet that are members of the partition.

5. The method of claim 4 , further comprising:
determining a set of the one or more switches to be updated based upon the generated spanning tree.

6. The method of claim 5 , further comprising:
distributing the generated spanning tree to each determined switch.

7. The method of claim 6 , further comprising:
updating, at each determined switch, a multicast forwarding table based upon the generated spanning tree.

8. A system for providing multicast group membership defined relative to partition membership in a high performance computing environment, comprising:
at a computer comprising one or more microprocessors;
a subnet, the subnet comprising one or more switches, a plurality of end nodes interconnected via the one or more switches, and a subnet manager running on a node of the subnet;
wherein a multicast join request message is generated at an end node of the subnet, wherein the multicast join message comprises a plurality of subnet administration attributes, wherein the end node is a member of a partition defined within the subnet, wherein a set of the plurality of end nodes, including the end node, are defined as members of the partition; and
wherein an indication is provided, via one of the plurality of subnet administration attributes of the multicast join message, that each of the set of the plurality of end nodes should be added to a same multicast group.

9. The system of claim 8 ,
wherein an indication is provided, via another of the plurality of subnet administration attributes of the multicast join message, of an identifier of the partition.

10. The system of claim 9 ,
wherein the multicast join request message is received at the subnet manager; and
wherein upon receiving the multicast join request message, all members of the partition are added to a same multicast group.

11. The system of claim 10 ,
wherein a spanning tree is generated, the spanning tree including all end nodes of the subnet that are members of the partition.

12. The system of claim 11 ,
wherein a determination is made as to a set of the one or more switches that are to be updated based upon the generated spanning tree.

13. The system of claim 12 ,
wherein the generated spanning tree is distributed to each determined switch.

14. The system of claim 13 ,
wherein a multicast forwarding table is updated at each determined switch, each update being based upon the generated spanning tree.

15. A non-transitory computer readable storage medium having instructions thereon for providing multicast group membership defined relative to partition membership in a high performance computing environment, which when read and executed cause a computer to perform steps comprising:
providing, at a computer comprising one or more microprocessors, a subnet, the subnet comprising one or more switches, a plurality of end nodes interconnected via the one or more switches, and a subnet manager running on a node of the subnet;
generating, at an end node of the subnet, a multicast join request message, wherein the multicast join message comprises a plurality of subnet administration attributes, wherein the end node is a member of a partition defined within the subnet, wherein a set of the plurality of end nodes, including the end node, are defined as members of the partition; and
indicating, via one of the plurality of subnet administration attributes of the multicast join message, that each of the set of the plurality of end nodes should be added to a same multicast group.

16. The non-transitory computer readable storage medium of claim 15 , the steps further comprising:
indicating, via another of the plurality of subnet administration attributes of the multicast join message, an identifier of the partition.

17. The non-transitory computer readable storage medium of claim 16 , the steps further comprising:
receiving, at the subnet manager, the multicast join request message; and
upon receiving the multicast join request message, adding all members of the partition to a same multicast group.

18. The non-transitory computer readable storage medium of claim 17 , the steps further comprising:
generating a spanning tree, the spanning tree including all end nodes of the subnet that are members of the partition.

19. The non-transitory computer readable storage medium of claim 18 , the steps further comprising:
determining a set of the one or more switches to be updated based upon the generated spanning tree.

20. The non-transitory computer readable storage medium of claim 19 , the steps further comprising:
distributing the generated spanning tree to each determined switch; and
updating, at each determined switch, a multicast forwarding table based upon the generated spanning tree.

Assignees

• Oracle Int Corp

Inventors

• Johnsen Bjørn Dag
• Bogdanski Bartosz
• Holen Line

Classifications

• H04L45/48

  Routing tree calculation · CPC title

• H04L45/484

  using multiple routing trees · CPC title

• H04L12/185 Primary

  with management of multicast group membership · CPC title

• G06F9/45558

  Hypervisor-specific management and integration aspects · CPC title

• G06F16/2255

  Hash tables · CPC title