You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I would expect that we would preserve the invariant that all hook pods have terminated (are in failed/succeeded/deleted) before we launch a new deployer pod, but that appears that is not the case today. What would we have to change to guarantee that?
The deployment controller can try to identify any preexisting deployers for the config of the deployment (before creating the new deployer) and delete if it finds any.
I would expect that we would preserve the invariant that all hook pods have terminated (are in failed/succeeded/deleted) before we launch a new deployer pod, but that appears that is not the case today. What would we have to change to guarantee that?
@ironcladlou @Kargakis
The text was updated successfully, but these errors were encountered: