praiskup / resalloc

Allocator and manager for (expensive) resources
GNU General Public License v2.0
7 stars 7 forks source link

More intransigent control for failed resources #87

Closed schlupov closed 2 years ago

schlupov commented 2 years ago

Fixes: https://pagure.io/copr/copr/issue/2083

Sometimes it happens that even though the resource has check_failed_count set to a number greater than 3, it is not deleted because it still has a ticket. This leads to endless jobs. We can be more intransigent and remove the resource after 10 checks, even though it still has a ticket.

schlupov commented 2 years ago

I don't know if it's too strict. Look into resalloc db on copr-be

select * from resources where id=952075;

sandboxed_since Saturday, April 23, 2022 3:57:14.716 AM

praiskup commented 2 years ago

I believe this is eventually resolved by #93. We'll re-consider merging this later if the problem happens again.

praiskup commented 2 years ago

Running for some time, and this problem seems to be resolved by #93. Anyway, thank you for the PR!