Conformance GKE: timeout reached waiting for NodePort #43749
Description
Conformance GKE is failing regularly on main with the following errors :
docker run --network host -v /tmp/tmp.LszEuAkFir/control-plane-kubeconfig.yaml:/root/.kube/config -v /home/runner/work/cilium/cilium:/root/app -v /home/runner/.aws:/root/.aws -v /home/runner/.azure:/root/.azure -v /home/runner/.config/gcloud:/root/.config/gcloud -e GITHUB_WORKFLOW_REF=cilium/cilium/.github/workflows/conformance-gke.yaml@refs/heads/main quay.io/cilium/cilium-cli-ci:cae79a0fc6c891288b3d7b145b2478b074a452ea cilium connectivity test --collect-sysdump-on-failure --log-code-owners --code-owners=CODEOWNERS --exclude-code-owners=@cilium/github-sec --external-target=google.com. --external-cidr=8.0.0.0/8 --external-ip=8.8.8.8 --external-other-ip=8.8.4.4 --hubble=false --flow-validation=disabled --test-concurrency=5 --junit-file 'cilium-junits/Installation and Connectivity Test (1.31, us-west3-a, 2, no-tunnel, 1).xml' --junit-property 'github_job_step=Run connectivity test (1.31, 1, no-tunnel)'
[...]
⌛ [gke_***_us-west3-a_cilium-cilium-20981427997-1-1] Waiting for NodePort 34.106.17.216:31639 (cilium-test-4/echo-other-node) to become ready...
⌛ [gke_***_us-west3-a_cilium-cilium-20981427997-1-1] Waiting for NodePort 34.106.17.216:30648 (cilium-test-5/echo-same-node) to become ready...
⌛ [gke_***_us-west3-a_cilium-cilium-20981427997-1-1] Waiting for NodePort 34.106.17.216:30447 (cilium-test-3/echo-other-node) to become ready...
⌛ [gke_***_us-west3-a_cilium-cilium-20981427997-1-1] Waiting for NodePort 34.106.158.207:30733 (cilium-test-1/echo-other-node) to become ready...
⌛ [gke_***_us-west3-a_cilium-cilium-20981427997-1-1] Waiting for NodePort 34.106.17.216:32640 (cilium-test-2/echo-other-node) to become ready...
timeout reached waiting for NodePort 34.106.17.216:31639 (cilium-test-4/echo-other-node) (last error: command failed (pod=cilium-test-4/client-b65598b6f-9g2dk, container=): context deadline exceeded)
timeout reached waiting for NodePort 34.106.17.216:30648 (cilium-test-5/echo-same-node) (last error: command failed (pod=cilium-test-5/client-b65598b6f-5dnjh, container=): context deadline exceeded)
timeout reached waiting for NodePort 34.106.17.216:30447 (cilium-test-3/echo-other-node) (last error: command failed (pod=cilium-test-3/client3-75555c5f5-cqsft, container=): context deadline exceeded)
timeout reached waiting for NodePort 34.106.158.207:30733 (cilium-test-1/echo-other-node) (last error: command failed (pod=cilium-test-1/client-b65598b6f-tsw4m, container=): context deadline exceeded)
timeout reached waiting for NodePort 34.106.17.216:32640 (cilium-test-2/echo-other-node) (last error: command failed (pod=cilium-test-2/client-b65598b6f-p4j45, container=): context deadline exceeded)
Example: https://github.com/cilium/cilium/actions/runs/20981427997/job/60307006134
Of interest: the config tunnel-ingress-controller doesn't seem to be affected. It's the only config that has KPR enabled.
Metadata
Metadata
Assignees
Labels
Continuous Integration testing issue or flakeImpacts bpf/ or low-level forwarding details, including map management and monitor messages.Issues related to kube-proxy (not the kube-proxy-free mode).This is a known failure that occurs in the tree. Please investigate me!Related to integration with cloud environments such as AKS, EKS, GKE, etc.