startupProbe being ignored after restarting the pod

[judab5ericom]

/sig architecture
/wg K8s Infra

What happened:

Kubernetes bring up Container successfully perform StartupProbe properly,
a few minutes later I make the POD restart by fail the Liveness Prob.
The container goes down and in the next creation of the POD the startupProbe does not run at all,
the container starts directly with Liveness and Readiness while startupProbe skipped

What you expected to happen:

I expect that startup probe will run every POD creation

How to reproduce it (as minimally and precisely as possible):

Configure POD with all three probes: Startup/Liveness/Readiness Probe (Probe should configure to run as EXEC shell script)
Verify that StartupProbe pass sucessfully, Verify that Liveness and Readiness pass sucessfully
Wait few minutes
Make Liveness to fail
Verify that POD restart due to Liveness Check,
Verify that POD go up again
Check Logs you should see Liveness probe without startup probe

Anything else we need to know?:

Environment:

• Kubernetes version (use kubectl version):
  Client Version: version.Info{Major:"1", Minor:"20", GitVersion:"v1.20.4", GitCommit:"e87da0bd6e03ec3fea7933c4b5263d151aafd07c", GitTreeState:"clean", BuildDate:"2021-02-18T16:12:00Z", GoVersion:"go1.15.8", Compiler:"gc", Platform:"linux/amd64"}
  Server Version: version.Info{Major:"1", Minor:"18", GitVersion:"v1.18.18", GitCommit:"6f6ce59dc8fefde25a3ba0ef0047f4ec6662ef24", GitTreeState:"clean", BuildDate:"2021-04-15T03:23:41Z", GoVersion:"go1.13.15", Compiler:"gc", Platform:"linux/amd64"}

YAML Code Snip:

• Cloud provider or hardware configuration:
  Instance run with Ubuntu
  $ cat /etc/issue
  Ubuntu 18.04.5 LTS \n \l

Using Rancher Docker Kubernetes

• OS (e.g: cat /etc/os-release):

• Kernel (e.g. uname -a):
  Linux xxx 5.4.0-1044-oracle Some documentation tweaks #47~18.04.1-Ubuntu SMP Thu Apr 22 03:29:37 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux

• Install tools:
  Rancher

• Network plugin and version (if this is a network-related bug):

• Others:
  YAML
  startupProbe:
  exec:
  command:
  - /startup_probe.sh
  failureThreshold: 30
  periodSeconds: 10
  successThreshold: 1
  timeoutSeconds: 1

livenessProbe:
exec:
command:
- /liveness.sh
failureThreshold: 30
periodSeconds: 10
successThreshold: 1
timeoutSeconds: 5

   readinessProbe:
      exec:
        command:
        - /readiness.sh
      failureThreshold: 3
      periodSeconds: 10
      successThreshold: 1
      timeoutSeconds: 5

[judab5ericom]

/sig architecture
/wg K8s Infra

[k8s-ci-robot]

@judab5ericom: The label(s) wg/k8s, wg/infra cannot be applied, because the repository doesn't have them.

In response to this:

/sig architecture
/wg K8s Infra

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[odinuge]

/cc @matthyx

[matthyx]

/assign
Will take a look on Tuesday... I think this was fixed on master but not sure about cherry picks.
I'll search once back in front of a real computer...

[matthyx]

/sig node

[wzshiming]

This is a duplicate Issue #101064
The affected versions are 1.18.18, 1.19.10, 1.20.6, 1.21.0, please try another version.

[pacoxu]

@wzshiming it shoud be fixed in 1.18.19 and 1.18.20 by #101226.
/close

Feel free to reopen if it still exists. @judab5ericom

[k8s-ci-robot]

@pacoxu: Closing this issue.

In response to this:

@wzshiming it shoud be fixed in 1.18.19 and 1.18.20 by #101226.
/close

Feel free to reopen if it still exists. @judab5ericom

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[pacoxu]

/remove-sig architecture
/triage accepted