Change Details

Multiple failing CI jobs that use catalyst can cause an overload of pods which stop new deployments from being scheduled. This affects patchdemo users and blocks them from creating bespoke demos. Let's use a separate node for ci jobs, so that when it runs out of space it won't affect patchdemo users. Developers can free up space for their CI runs by deleting wikis belonging to failed jobs if needed (these get cleaned up automatically in 3 days).

Multiple failing CI jobs that use catalyst can cause an overload of pods which stop new deployments from being scheduled. This affects patchdemo users and blocks them from creating bespoke demos. AC - [x] Check in with Abstract Wikipedia folks to see if they need these environments - [x] If not, ensure they are automatically reaped just like successful test runs

Multiple failing CI jobs that use catalyst can cause an overload of pods which stop new deployments from being scheduled. This affects patchdemo users and blocks them from creating bespoke demos. Let's use a separate node for ci jobs, so that when it runs out of space it won't affect patchdemo users. AC - [x] Check in with Abstract Wikipedia folks to see if they need these environments - [x] If not, Developers can free up space for their CI runs by deleting wikis belonging to failed jobs if needed (these get cleaned up automatically in 3 days).ensure they are automatically reaped just like successful test runs