System and method for supporting node role attributes in a high performance computing environment
System and method for supporting node role attributes in a high performance computing environment. In accordance with an embodiment, a node role attribute can comprise a vendor defined subnet management attribute. When a subnet manager attempts to discover a high performance computing environment, such as an InfiniBand subnet, or a switch topology, identifying a topology is quite complex when subnet manager can only observe connectivity, without context behind the connectivity (the roles of the different nodes in the connectivity). However, when a subnet has a node role attribute enabled, the subnet manager can map the interconnect more effectively as it can discover not only the connectivity during the initial sweep, but it can also discover the role of each node discovered, thus leading to a more efficient interconnect discovery.
1. A system for supporting node role attributes in a high performance computing environment, comprising:
one or more microprocessors; and
a subnet, the subnet comprising:
a plurality of switches, the plurality of switches being arranged in a Fat-Tree topology,
a management switch, the management switch having global connectivity within the subnet, and
a subnet manager, the subnet manager running on one of the plurality of switches, wherein the subnet manager, on startup of the subnet, performs a discovery sweep and a routing of the subnet;
wherein the management switch is assigned a node role attribute of a plurality of node role attributes, wherein the node role attribute of the management switch comprises a management switch node role attribute;
wherein the subnet manager, prior to performing the discovery sweep of the subnet, queries the node role attribute of the management switch; and
wherein the subnet manager, upon discovery of the management switch node role attribute at the management switch, excludes the management switch from the routing of the subnet.
2. The system of claim 1 ,
wherein the plurality of switches comprises a leaf switch;
wherein the leaf switch is assigned another node role attribute of a plurality of node role attributes, the node role attribute assigned to the leaf switch comprising a leaf switch node role attribute.
3. The system of claim 2 ,
wherein the subnet manager, prior to performing the discovery sweep of the subnet, queries the node role attribute of the leaf switch;
wherein the subnet manager, upon discovery of the leaf switch node role attribute at the leaf switch, utilizes the role of the leaf switch in routing the subnet.
4. The system of claim 3 ,
wherein the subnet manager stores an association between the leaf switch and the leaf switch node role attribute at a memory accessible by the subnet manager.
5. The system of claim 4 ,
wherein the subnet undergoes a reconfiguration triggering event, the reconfiguration triggering event causing the subnet manager to perform a re-routing of the subnet; and
wherein during the re-routing of the subnet, the subnet manager utilizes the stored association between the leaf switch and the leaf switch node role attribute of the leaf switch in re-routing the subnet.
6. The system of claim 5 ,
wherein the reconfiguration triggering event comprises a topology change of the subnet.
7. A method for supporting node role attributes in a high performance computing environment, comprising:
providing, at one or more computers, including one or more microprocessors, a subnet, the subnet comprising:
a plurality of switches, the plurality of switches being arranged in a Fat-Tree topology,
a management switch, the management switch having global connectivity within the subnet, and
a subnet manager, the subnet manager running on one of the plurality of switches, wherein the subnet manager, on startup of the subnet, performs a discovery sweep and a routing of the subnet;
assigning the management switch a node role attribute of a plurality of node role attributes, wherein the node role attribute of the management switch comprises a management switch node role attribute;
querying, by the subnet manager, prior to performing the discovery sweep of the subnet, the node role attribute of the management switch;
excluding, by the subnet manager, upon discovery of the management switch node role attribute at the management switch, the management switch from the routing of the subnet.
8. The method of claim 7 ,
wherein the plurality of switches comprises a leaf switch;
wherein the leaf switch is assigned another node role attribute of a plurality of node role attributes, the node role attribute assigned to the leaf switch comprising a leaf switch node role attribute.
9. The method of claim 8 ,
wherein the subnet manager, prior to performing the discovery sweep of the subnet, queries the node role attribute of the leaf switch;
wherein the subnet manager, upon discovery of the leaf switch node role attribute at the leaf switch, utilizes the role of the leaf switch in routing the subnet.
10. The method of claim 9 ,
wherein the subnet manager stores an association between the leaf switch and the leaf switch node role attribute at a memory accessible by the subnet manager.
11. The method of claim 10 ,
wherein the subnet undergoes a reconfiguration triggering event, the reconfiguration triggering event causing the subnet manager to perform a re-routing of the subnet; and
wherein during the re-routing of the subnet, the subnet manager utilizes the stored association between the leaf switch and the leaf switch node role attribute of the leaf switch in re-routing the subnet.
12. The method of claim 11 ,
wherein the reconfiguration triggering event comprises a topology change of the subnet.
13. A non-transitory computer readable storage medium, including instructions stored thereon for supporting node role attributes in a high performance computing environment, which when read and executed by one or more computers cause the one or more computers to perform steps comprising:
providing, at one or more computers, including one or more microprocessors, a subnet, the subnet comprising:
a plurality of switches, the plurality of switches being arranged in a Fat-Tree topology,
a management switch, the management switch having global connectivity within the subnet, and
a subnet manager, the subnet manager running on one of the plurality of switches, wherein the subnet manager, on startup of the subnet, performs a discovery sweep and a routing of the subnet;
assigning the management switch a node role attribute of a plurality of node role attributes, wherein the node role attribute of the management switch comprises a management switch node role attribute;
querying, by the subnet manager, prior to performing the discovery sweep of the subnet, the node role attribute of the management switch;
excluding, by the subnet manager, upon discovery of the management switch node role attribute at the management switch, the management switch from the routing of the subnet.
14. The non-transitory computer readable storage medium of claim 13 ,
wherein the plurality of switches comprises a leaf switch;
wherein the leaf switch is assigned another node role attribute of a plurality of node role attributes, the node role attribute assigned to the leaf switch comprising a leaf switch node role attribute.
15. The non-transitory computer readable storage medium of claim 14 ,
wherein the subnet manager, prior to performing the discovery sweep of the subnet, queries the node role attribute of the leaf switch;
wherein the subnet manager, upon discovery of the leaf switch node role attribute at the leaf switch, utilizes the role of the leaf switch in routing the subnet.
16. The non-transitory computer readable storage medium of claim 15 ,
wherein the subnet manager stores an association between the leaf switch and the leaf switch node role attribute at a memory accessible by the subnet manager.
17. The non-transitory computer readable storage medium of claim 16 ,
wherein the subnet undergoes a reconfiguration triggering event, the reconfiguration triggering event causing the subnet manager to perform a re-routing of the subnet;
wherein during the re-routing of the subnet, the subnet manager utilizes the stored association between the leaf switch and the leaf switch node role attribute of the leaf switch in re-routing the subnet; and
wherein the reconfiguration triggering event comprises a topology change of the subnet.