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
As discussed/noted in this thread. The static upgrade timeout of 15 minutes is likely too short for production usage.
Before blindingly changing the timeout we should do some investigating to understand why upgrades take so long (This reportedly surfaces most frequently in clusters under heavy load).
The timeout (very likely) should scale linearly with the total number of nodes rather than being a static value.
As discussed/noted in this thread. The static upgrade timeout of 15 minutes is likely too short for production usage.
Before blindingly changing the timeout we should do some investigating to understand why upgrades take so long (This reportedly surfaces most frequently in clusters under heavy load).
The timeout (very likely) should scale linearly with the total number of nodes rather than being a static value.
See also:
rpk cluster health
should not be used in a readiness probe helm-charts#1002JIRA Link: K8S-341
The text was updated successfully, but these errors were encountered: