I never upgraded from 1.13.1 to 1.14.0, i went straight to 1.14.1 using the changes of this pull request.
I do want to highlight that it didn't go as smooth as I'd like. Everything went as expected until the end of the playbook. My NGINX ingress controllers all went down for a short period. I have a hard time figuring out why.
I see these events in my NGINX ingress controller pods:
Received SIGTERM, shutting down
k8s.io/ingress-nginx/internal/ingress/controller/store/store.go:153: watch of *v1.Service ended with: too old resource version: 18603171 (18603183)
I believe the SIGTERM comes from the readiness probe returning 500, but I can't find any indications for why it would return 500. :(
I don't believe this is related to 1.14.1 though, this would probably have happened if I went to 1.14.0 too, so I think we can ignore that when we review this pull request. It would be nice to know what happened though.
I never upgraded from
1.13.1
to1.14.0
, i went straight to1.14.1
using the changes of this pull request.I do want to highlight that it didn't go as smooth as I'd like. Everything went as expected until the end of the playbook. My NGINX ingress controllers all went down for a short period. I have a hard time figuring out why.
I see these events in my NGINX ingress controller pods:
And I see logs like this from the pods:
I believe the
SIGTERM
comes from the readiness probe returning 500, but I can't find any indications for why it would return 500. :(I don't believe this is related to
1.14.1
though, this would probably have happened if I went to1.14.0
too, so I think we can ignore that when we review this pull request. It would be nice to know what happened though.