HP-UX Workload Manager A.02.03.01 Release Notes for HP-UX 11i V1.0 and HP-UX 11i V2.0

HP-UX Workload Manager Release Notes

Known problems and workarounds

Temporary iCOD capacity expires while WLM is managing

nPartitions

Issue

WLM manages nPartitions using its wlmpard daemon. Assume wlmpard is

started on a system that has temporary iCOD capacity in use. If that

temporary capacity expires, wlmpard will still be able to deactivate CPUs

without any problems. However, wlmpard may attempt to activate CPUs

based on the expired capacity. These attempts will fail because the

temporary capacity no longer exists. wlmpard will not abort, but it may

continue to attempt to activate unavailable CPUs, generating a message of

the following form in /var/opt/wlm/msglog:

Error increasing CPU count on partition par_name (has x needs

y).

Workaround

Stop wlmpard (using the -k option) and restart it so that it is aware of the

currently licensed capacity on the system.

CPUs are not being released

Issue

When managing either:

• Partitions, or

• PSET-based workload groups with SLOs

groups can have artificially high CPU allocations as the result of another

SLO or the distribution of excess CPU resources. Under WLM’s default

allocation algorithm, the groups may be slow, or even unwilling, to release

unneeded CPU resources.

Workaround

To ensure groups release unneeded CPU resources, set the

cntl_base_previous_req tunable to 1. For more information, see the

wlmconf(4) man page.