update node controller to use shared pod informer#24841
update node controller to use shared pod informer#24841k8s-github-robot merged 2 commits intokubernetes:masterfrom
Conversation
|
cc @kubernetes/rh-cluster-infra @smarterclayton @deads2k @liggitt |
635174e to
346a782
Compare
|
This PR is not for the master branch but does not have the |
@eparis I feel like the bot is trying to tell me something, but I don't know what. |
|
Its trying to tell you the github itself had major problems and the bot lost it's mind. |
|
not sure who is the best reviewer -- @lavalamp ? @wojtek-t ? @bprashanth ? Assigning to @wojtek-t for now. |
06c512b to
f7a3d11
Compare
There was a problem hiding this comment.
Do not update the deployment controller to share the pod cache just yet because it mutates its pod cache! I've opened #27687
|
@mwielgus FYI, as this intersects with some stuff he's doing on Federation. |
|
@quinton-hoole - I don't think so; this is something different |
f7a3d11 to
db6026c
Compare
|
Ok finally updated this. Dropped the deployment controller (see above why), rebased, refactored a change that introduced an incompatibility with the node controller using the shared informer. |
|
@deads2k is the cache mutation detector enabled for Kube? |
|
@Kargakis what is "cache mutation detector"? |
It allows detection of cache mutations based on an env var. It's not cheap to run, but its useful when running e2e style tests to find people who are messing with the cache. I'd really like to see it (or something that does the same job) merged. |
|
LGTM |
|
GCE e2e build/test passed for commit db6026c. |
|
@k8s-bot test this [submit-queue is verifying that this PR is safe to merge] |
|
GCE e2e build/test passed for commit db6026c. |
|
Automatic merge from submit-queue |
[release-4.3] Bug 1821341: UPSTREAM: 88440: Report deleted pod status correctly Origin-commit: 21a4b974e162e4a2acfa77a03cf2e10986f12d48
continuing work from #24470 and #23575