Mark Grondona
Mark Grondona
If we could get the resource module to update its version of R, then we could sidestep the resource acquisition protocol for now by reloading the scheduler. Then we could...
> Edit: oh except feasibility check wouldn't work. It would be easy to write a more generic feasibility validator plugin that rejected obviously infeasible jobs in this case if it...
Actually, in the case of a DAT the eventual resources are probably already known, so maybe this is a non-issue for that use case.
> Hey for the DAT case, could we remove resources.R from the KVS, then reload then resource module across the (full) instance and expect it to build a new resource.R...
For completeness, another idea discussed was to add a separate service that would collect resource information from the `resource` module and the `job-manager` and give consumers a single endpoint to...
I think this is due to known issue #4108. There is no `epilog-finish` event because the perilog plugin lost track of the epilog process after the restart. A solution for...
Here's a jobtap plugin that might work to post the missing `epilog-finish` events for jobs in this state. It should be loaded like: ```console $ flux jobtap load /path/to/plugin.so jobs="[ID1,...
FYI, I just ran the above jobtap plugin against all CLEANUP jobs on Corona and they're all inactive now, e.g.: ``` # flux job eventlog f2cx7pwPP9CK 1700501129.045566 submit userid=60943 urgency=16...
Probably not, that was just addition of a utility to fix this state. We can consider this resolved when we can recapture or restart any pending epilog processes/procedures and ensure...