Distributed subnet manager for infiniband networks

摘要

In an approach for distributed management of a network, a processor discovers a network, wherein the network is comprised of a plurality of nodes and an at least one switch. A processor identifies a first port associated with the at least one node. A processor identifies a link between the first port and a second port within the network. A processor determines the at least one link is properly configured. A processor selects a first node of the plurality of nodes to be a master node, wherein the master node is configured to manage the network. A processor initializes a routing table, wherein the routing table contains locations of each node in the plurality of nodes and a location of the at least one switch. A processor adjusts by the first node, the routing table based on the change to the network.

主权项

1. A method for distributed management of a network, the method comprising:
discovering, by one or more processors, a network, wherein the network is comprised of a plurality of nodes and at least one switch; identifying, by one or more processors, a first port associated with the at least one node; identifying, by one or more processors, a link between the first port and a second port within the network; determining, by one or more processors, the at least one link is properly configured; selecting, by one or more processors, a first node of the plurality of nodes, wherein the first node provides subnet administrator functionality; determining, by one or more processors, that the first node is not a master node, wherein the master node is configured to manage the network; identifying, by one or more processors, a second node that is the master node; determining, by one or more processors, that the second node has entered a standby mode; selecting, by one or more processors, the first node to be the master node; initializing, by one or more processors, a routing table, wherein the routing table contains locations of each node of the plurality of nodes and a location of the at least one switch; identifying, by one or more processors, a change to the network; identifying, by one or more processors, that a third node within the network is malfunctioning; establishing, by one or more processors, a routing path, wherein the routing path is a cross node routing path; identifying, by one or more processors, that a fourth node is non-responding; resetting, by one or more processors, a port associated with the fourth node; assigning, by one or more processors, a network address to each node incorporated into the network; invalidating, by one or more processors, network addresses associated with each node removed from the network; determining, by one or more processors, that an address associated with at least one port has changed; invalidating, by one or more processors, the address associated with the at least one port; identifying, by one or more processors, an updated address associated with the at least one port; and adjusting, by one or more processors, the routing table to include the updated address associated with the at least one port, based on the change to the network.