Extended.[k8s.io] Probing container should *not* be restarted with a /healthz http liveness probe [Conformance] #12072
Comments
bparees
added
component/kubernetes
kind/test-flake
|
Same as kubernetes/kubernetes#28084 |
|
I disabled this on origin_gce to due to flaking |
|
It got noted again in upstream via kubernetes/kubernetes#30714 Saw it on our end again in #13577 |
|
referenced issue above has been closed so i am raising the priority of this issue. |
|
This test sucks, I think there may be a serious problem here. Has flaked
for last two years.
On Oct 9, 2017, at 6:54 PM, Ben Parees <[email protected]> wrote:
https://openshift-gce-devel.appspot.com/build/origin-ci-test/pr-logs/pull/16758/test_pull_request_origin_extended_conformance_install_update/7582/
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#12072 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ABG_p99dqMmIVVOYzt2AkJpKJMGJDm6tks5sqqRAgaJpZM4K_zvJ>
.
|
|
I have been trying to reproduce this locally but without luck. The main idea of this test is to check that a good container won't restart. Looking at the code and the history of this issue, following are my observations.
Since the goal this test is to check that a good container won't restart based on liveness probe, would it make sense to add a liveness command that checks for the existence of a particular directory or something else instead of a HTTP GET. |
|
@ravisantoshgudimetla do you feel like you can deliver the change to the liveness probe to move away from HTTP? |
|
Flaked again-https://ci.openshift.redhat.com/jenkins/job/test_pull_request_origin_extended_conformance_gce/15413/console @stevekuznetsov - I created an upstream issue yesterday. I think it would be better to delete this test. I did not get any feedback on it yet but I will create a PR and see if the upstream is ok with it. |
Automatic merge from submit-queue (batch tested with PRs 60342, 60505, 59218, 52900, 60486). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>. Increase failureThresholds for failing HTTP liveness test **What this PR does / why we need it**: Removes test from e2e which relies on HTTP liveness as a measure to tell if the container is good or bad. While this is not a bad idea, we cannot rely on this test as HTTP liveness relies on network/infrastructure etc on which sometimes we have no control over. While increasing the timeout may be an option it may not be ideal for all cloud providers/type of hardware etc. **Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*: Fixes #59150 **Special notes for your reviewer**: I have stated reasons in the issue #59150. We have seen that this test is flaking recently in openshift/origin#12072 **Release note**: ```release-note NONE ```
|
Issues go stale after 90d of inactivity. Mark the issue as fresh by commenting If this issue is safe to close now please do so with /lifecycle stale |
openshift-ci-robot
added
the
lifecycle/stale
|
Stale issues rot after 30d of inactivity. Mark the issue as fresh by commenting If this issue is safe to close now please do so with /lifecycle rotten |
openshift-ci-robot
added
lifecycle/rotten
|
Rotten issues close after 30d of inactivity. Reopen the issue by commenting /close |
https://ci.openshift.redhat.com/jenkins/job/test_pull_requests_origin_conformance/8991/
The text was updated successfully, but these errors were encountered: