awx
awx copied to clipboard
RFE: Resource limit for job or fork.
ISSUE TYPE
- Feature Idea
SUMMARY
Add configurable resource limits for jobs or forks. If the job reaches the limit during executions, kill the fork or job itself to protect other jobs running on the same Ansible Tower instance. If the limits were to be set on the jobs or forks, this will also help the scheduler in addition to the capacity to determine on which instance to run the job on.
This is not something we currently have control over in general.
Pod specs for container group do allow for some of this.
I think we could probably do this in the future with jobs directly with podman... I don't think I'd like the UX around killing the job specifically. Rather just setting upper limits on how much resources can be consumed... the processes inside of the pods could decide to die or bog down.
Any update on this with respect to AAP 2.x?