...
In order to support Cray features like requesting particular NUMA nodes, or some number of NUMA nodes, PBS had created one vnode for each NUMA node.
Sites have requested the ability to change the way PBS reports the Cray nodes, so that they can have fewer vnodes to deal with and not
encounter some of the differences that multiple vnodes per host brings.
PBS shall provide the option to create a vnode per NUMA node or a vnode per compute node on a Cray system.
The default behavior of PBS will be changed to create a vnode per compute node by default.
PBS shall continue to support the backwards compatible behavior of creating one vnode per NUMA node.
Interface: vnode_per_computenuma_node
- Visibility: Public PBS Private
- Change Control: Experimental
- Details:
- New interface: new mom_priv/config variable.
- alps_create_vnode_per_numa is a Boolean. Setting it to "TRUE" will cause PBS to create one vnode per Cray numa node. This is the
default behavior when alps_create_vnode
Setting it to "FALSE" will cause PBS to create one vnode per Cray compute node. This is the default behavior
when $vnode_per_numa_node is not set in mom_priv/config. Setting it to "FALSE" will cause PBS to create
one vnode per Cray compute node. - Note: when alps$vnode_create_vnode_per_numa is set to "FALSE", it will not be possible for PBS to schedule based on segments.
...
Administrator's instructions.
- When changing from one version of PBS using one type of vnode create versus another, you must first:
- Quiesce the system (ensure there are no running jobs).
- Remove all existing vnodes of the vntype cray_compute. The easiest way to do this is to delete all vnodes using
- qmgr -c "delete node @default"
- Set the mom_priv/config variable to desired setting for all Moms.
- HUP or restart the MoM.
- Add Stop the version of the mom that create vnodes per NUMA node
- Start the new version of MoM that will create vnodes per compute node
- All moms on a Cray X* system must be of the same PBS version.
- On the server add the MoMs back using qmgr: qmgr -c "create node <mom name>"
...