Skip to content

[OCPEDGE-2381] OCPBUGS-83782: Add test to verify for backup container exists when etcd crashes #30922

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
openshift-merge-bot merged 1 commit into from
Apr 16, 2026
+61 −0
Merged

[OCPEDGE-2381] OCPBUGS-83782: Add test to verify for backup container exists when etcd crashes #30922

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
61 changes: 61 additions & 0 deletions test/extended/two_node/tnf_recovery.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@ import (
"fmt"
"math/rand"
"os"
"strings"
"time"

g "github.com/onsi/ginkgo/v2"
Expand Down Expand Up @@ -407,6 +408,66 @@ var _ = g.Describe("[sig-etcd][apigroup:config.openshift.io][OCPFeatureGate:Dual
&targetNode, true, false, // targetNode expected started == true, learner == false
6*time.Minute, 45*time.Second)
})

g.It("should leave a backup container behind for debugging when etcd container crashes", func() {
survivedNode := peerNode

g.By("Recording epoch timestamp before reboot")
epochStr, err := exutil.DebugNodeRetryWithOptionsAndChroot(oc, targetNode.Name, "openshift-etcd",
"bash", "-c", "date +%s")
o.Expect(err).To(o.BeNil())
rebootEpoch := strings.TrimSpace(epochStr)

Comment thread
kasturinarra marked this conversation as resolved.
g.By(fmt.Sprintf("Cleaning up any stale etcd-previous container on %s", targetNode.Name))
_, _ = exutil.DebugNodeRetryWithOptionsAndChroot(oc, targetNode.Name, "openshift-etcd",
"bash", "-c", "podman rm -f etcd-previous 2>/dev/null || true")

g.By(fmt.Sprintf("Removing /var/lib/etcd/pod.yaml on %s", targetNode.Name))
_, err = exutil.DebugNodeRetryWithOptionsAndChroot(oc, targetNode.Name, "openshift-etcd",
"bash", "-c", "rm -f /var/lib/etcd/pod.yaml")
o.Expect(err).To(o.BeNil(), "Expected to remove pod.yaml without error")
Comment thread
coderabbitai[bot] marked this conversation as resolved.

g.By(fmt.Sprintf("Rebooting %s ungracefully", targetNode.Name))
err = exutil.TriggerNodeRebootUngraceful(oc.KubeClient(), targetNode.Name)
o.Expect(err).To(o.BeNil(), "Expected to trigger ungraceful reboot without error")
time.Sleep(time.Minute)

Comment thread
kasturinarra marked this conversation as resolved.
g.By(fmt.Sprintf("Ensuring that %s added %s back as learner (timeout: %v)", survivedNode.Name, targetNode.Name, memberIsLeaderTimeout))
validateEtcdRecoveryState(oc, etcdClientFactory,
&survivedNode,
&targetNode, false, true,
memberIsLeaderTimeout, utils.FiveSecondPollInterval)

g.By(fmt.Sprintf("Ensuring %s rejoins as learner (timeout: %v)", targetNode.Name, memberRejoinedLearnerTimeout))
validateEtcdRecoveryState(oc, etcdClientFactory,
&survivedNode,
&targetNode, true, true,
memberRejoinedLearnerTimeout, utils.FiveSecondPollInterval)

g.By(fmt.Sprintf("Ensuring %s node is promoted back as voting member (timeout: %v)", targetNode.Name, memberPromotedVotingTimeout))
validateEtcdRecoveryState(oc, etcdClientFactory,
&survivedNode,
&targetNode, true, false,
memberPromotedVotingTimeout, utils.FiveSecondPollInterval)

g.By(fmt.Sprintf("Verifying etcd container is running on %s", targetNode.Name))
got, err := exutil.DebugNodeRetryWithOptionsAndChroot(oc, targetNode.Name, "openshift-etcd",
strings.Split(ensurePodmanEtcdContainerIsRunning, " ")...)
o.Expect(err).To(o.BeNil())
o.Expect(got).To(o.Equal("'true'"), fmt.Sprintf("expected etcd container running on %s", targetNode.Name))

g.By(fmt.Sprintf("Verifying etcd-previous container exists on %s", targetNode.Name))
prevOutput, err := exutil.DebugNodeRetryWithOptionsAndChroot(oc, targetNode.Name, "openshift-etcd",
"bash", "-c", "podman ps -a --format '{{.Names}}' | grep -m1 etcd-previous")
o.Expect(err).To(o.BeNil(), fmt.Sprintf("expected etcd-previous container to exist on %s", targetNode.Name))
o.Expect(strings.TrimSpace(prevOutput)).To(o.Equal("etcd-previous"),
fmt.Sprintf("expected etcd-previous container on %s", targetNode.Name))

g.By(fmt.Sprintf("Verifying pod.yaml was recreated on %s via pacemaker log", targetNode.Name))
_, err = exutil.DebugNodeRetryWithOptionsAndChroot(oc, targetNode.Name, "openshift-etcd",
"bash", "-c", fmt.Sprintf("journalctl -u pacemaker --since=@%s --no-pager | grep -m1 -i 'a new working copy of /etc/kubernetes/static-pod-resources/etcd-certs/configmaps/external-etcd-pod/pod.yaml was created'", rebootEpoch))
o.Expect(err).To(o.BeNil(), "Expected pacemaker log to contain pod.yaml recreation entry after reboot")
})
})

func validateEtcdRecoveryState(
Expand Down