[release-v1.42] Use pod-network k8s service endpoint for non-cluster-host Typha#4842
Merged
caseydavenport merged 1 commit intoMay 21, 2026
Conversation
…ra#4840) The non-cluster-host Typha deployment runs pod-networked, but it was inheriting KUBERNETES_SERVICE_HOST/PORT from the host-network endpoint. On MKE clusters that resolves to proxy.local:6444, which is unreachable from pods, so the deployment crashloops with DNS timeouts. Override those env vars from K8sServiceEpPodNetwork (populated from the kubernetes-service-endpoint ConfigMap's *_POD_NETWORK keys) when set. (cherry picked from commit e3664ff)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Cherry-pick of #4840 to
release-v1.42(Calico Enterprise v3.23.x). Requested for inclusion in the v3.23.1 patch release.The non-cluster-host Typha deployment runs pod-networked, but it inherits its kube-apiserver env vars from the host-network endpoint. On MKE clusters that's
proxy.local:6444, which pods can't resolve, so the NCH Typha crashloops with DNS timeouts during the iptables-to-eBPF migration.Clean cherry-pick.
Note: the master follow-up #4846 (always-strip-then-readd) needs to be picked separately once it merges.
Related: https://tigera.atlassian.net/browse/CI-1987