Compute Nodes
Compute nodes are the nodes on which work runs in your IBM® Spectrum Cluster Foundation Community Edition system.
Every compute node has a location attribute. The node's location
information depends on the server type:
- For rack servers, the node's location information depends on which rack the node is in, the height of the node (in U) and the node's position in the rack.
- For blade servers or Flex System nodes, the node's location information depends on which rack and chassis the node is in.
A node's location information can be specified in the following
ways:
- A node's location is specified while nodes, that are specified
in the node information file, are imported. In this case, the node's
location information is created together with the node. In the node information file you can specify the following attributes:
- For rack servers: specify the rack name, node position, and node height.
- For blade servers or Flex System nodes: specify the rack name, and chassis name.
- A node's location is specified during the node discovery process.
In this case, the node's location information is created while the
nodes are discovered. The following attributes are generated automatically during the node discovery process:
- For rack servers, the rack name, and node height is generated incrementally during the node discovery process.
- For blade servers or Flex System nodes, the rack name, and chassis name is generated during the node discovery process.
- A node's location can be changed when you relocate a node. If node's location information is incorrect, the location can be reset by using the Relocate option in the Web Portal.