Closed pkking closed 1 year ago
This typically doesn't happen. Seems like the background worker was killed and it did not close the ticket? The answer might or might not be in worker.log, depending if it was killed by OOM or otherwise.
Huh, now I realized after the PTO that this one si really old. Likely non-issue. Sorry for the long delay.
Note there's the backend/run/copr-backend-unknown-resalloc-tickets.py script that might help to identify the dangling tickets (we have sometimes such tickets, but they are typically taken manually for debugging purposes).
Ah! This should have been reported against fedora-copr/copr.
This week we start some builds( about 2000 build) maybe a bit large. After the weekend we noticed that some tickets were not closed(no idea why) which cause new vm ticket can't be alloced.
For a workaround i close some ticket manually using
resalloc ticket-close <id>
Im still digging into to find the root cause, any suggestions appreciated :P