PBS should better handle cpuset MOM resource availability that persistently does not match actual cpuset resource availability resulting in cpuset creation failures

Description

The jobs get killed instantly due to cpuset creation failure, this however gets resolved if the PBS/mom services are restarted. But this would be a problem when they move to production.

Acceptance Criteria

None

Status

Assignee

Unassigned

Reporter

Former user

Severity

None

OS

None

Start Date

None

Pull Request URL

None

Components

Priority

Critical
Configure