USHIFT-1199: add automated tests for etcd memory limit management#1735
Conversation
|
@dhellmann: This pull request references USHIFT-1199 which is a valid jira issue. DetailsIn response to this: 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. |
|
/hold These tests depend on #1734 These tests introduce instability to the overall test run by restarting microshift and not waiting long enough for it to restart. |
9c5902e to
9a45753
Compare
9a45753 to
44da87a
Compare
281b088 to
d789ebc
Compare
d789ebc to
fece1d7
Compare
|
/assign @pacevedom |
0b445dd to
151dba5
Compare
151dba5 to
4aeb19d
Compare
9547aab to
15ec680
Compare
|
/retest-required |
|
/hold The e2e tests are not running on this PR because only test code is being changed. See openshift/release#39797 |
|
/retest-required |
|
/hold cancel |
|
/test microshift-e2e-arm |
1fb0962 to
32f6a74
Compare
|
@dhellmann: The following test failed, say
Full PR test history. Your PR dashboard. DetailsInstructions 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. I understand the commands that are listed here. |
32f6a74 to
fd4901a
Compare
There was a problem hiding this comment.
What do you think about cat /etc/microshift/config.yaml 2>/dev/null || true? This way we can check for rc==0 and avoid mistaking ssh errors for an empty configuration.
There was a problem hiding this comment.
If we always return true, we would have to look at the output to see if we got any content. Is that what you mean? I'm not sure when we would have an empty file.
There was a problem hiding this comment.
MicroShift runs as root so it will be able to read anything, isnt it?
There was a problem hiding this comment.
It does, but we want to make sure that file is owned by root for security and because that's what we would expect users to do.
There was a problem hiding this comment.
Actually it does work even if not in ostree. You will see an error about not being in an ostree host but thats only cosmetic. Checks run and return codes reflect situation. Needs some refactoring though, as it always expects topolvm to be up, even if there is no lvmd.
There was a problem hiding this comment.
I think I tried the script and it exited because I was not on an ostree system. Maybe I was running the wrong one.
There was a problem hiding this comment.
Greenboot tests do this in a different way, will ping you when its up for review. Actually it would benefit from some of the work in this PR.
There was a problem hiding this comment.
@pmtk has some "wait" logic in one of his PRs, too. It would be ideal if we could come up with 1 good strong implementation of that.
There was a problem hiding this comment.
This is already covered by sos report, isnt it? It doesnt even matter if the test succeeds or not (it fails here, though). You are going to get a huge output from this, mixed with the other tests/keywords output.
There was a problem hiding this comment.
We may want to cut this down, but I found it useful in debugging to see the logs from the process at the time of the command inline. I could make it 20 lines?
Signed-off-by: Doug Hellmann <dhellmann@redhat.com>
fd4901a to
8a0c986
Compare
pacevedom
left a comment
There was a problem hiding this comment.
Will address comments in further PRs, as the main target of the PR looks good. Merging for progress and lets keep iterating the common functions.
/lgtm
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: dhellmann, pacevedom The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
No description provided.