pkg
pkg copied to clipboard
Adopt health checks in sharedmain
Currently, we do not use health checks in our controllers descended from sharedmain
. Adopting health checks will allow us to begin reporting controllers as unhealthy so that they are restarted by the kubelet. For example, we will be able to use the leaderelection health checker to begin reporting unhealthy after the leader election timeout elapsed when renewing the leader election lease.
/kind feature
Expected Behavior
Controllers can report themselves as unhealthy so that they can be restarted.
Actual Behavior
That doesn't happen.
This issue is stale because it has been open for 90 days with no
activity. It will automatically close after 30 more days of
inactivity. Reopen the issue with /reopen
. Mark the issue as
fresh by adding the comment /remove-lifecycle stale
.
/open
/lifecycle frozen
We should use the shared drainer
logic, but probably have something that does this over HTTP (no need for TLS)
/assign
I'll take a shot at this.
Unassigned so someone can resolve this faster than I can.