Staging instance, all changes can be removed at any time

Skip to content

New priority proposal to avoid preemption of system/not swh pods with the default priority

Vincent Sellier requested to merge new_priorities_proposal into staging

Some components deployed via external helm charts can have a priority defined. They are evicted when the resources start to be short.

To avoid that, we can declare our low priority components with a priority < 0 so they will be preempted in favour of other cluster components like opentelemetry exporter or some operators

Only in staging for now to validate it's ok.

Merge request reports

Loading