Use a differently-named DC in test/cmd/idle.sh #12590
Conversation
This should "fix" a long-running issue involving flakes due to creation/deletion of DCs with the same name in said tests by making sure that the "old" DC won't accidentally interfere with the new DC. Partial for openshift#11004
|
[test] |
2 similar comments
|
[test] |
|
[test] |
|
Evaluated for origin test up to 0bf1beb |
|
continuous-integration/openshift-jenkins/test SUCCESS (https://ci.openshift.redhat.com/jenkins/job/test_pr_origin/13149/) (Base Commit: e478226) |
|
[merge] |
|
Evaluated for origin merge up to 0bf1beb |
|
continuous-integration/openshift-jenkins/merge SUCCESS (https://ci.openshift.redhat.com/jenkins/job/test_pr_origin/13156/) (Base Commit: 6e9d68d) (Image: devenv-rhel7_5748) |
|
Well, this apparently did not fix it. |
| os::cmd::expect_success 'oc create -f test/extended/testdata/idling-echo-server.yaml' | ||
| os::cmd::expect_success 'oc describe deploymentconfigs idling-echo' | ||
| os::cmd::expect_success 'oc create -f test/testdata/idling-svc-route.yaml' | ||
| dc_name=$(basename $(oc create -f test/testdata/idling-dc.yaml -o name)) # `basename type/name` --> name |
There was a problem hiding this comment.
I'm not particularly happy with this. In a follow-up we should either:
- pass it in through a template
- grab it using
jsonpathfrom the metadata
There was a problem hiding this comment.
-o name is the only supported flag here. We could grab the resource/name with -o name and then extract it using oc get $(resource_with_name) -o go-template=sometemplate or somesuch, but that seems a bit redundant.
Passing it through a template seems like a decent idea, though.
This should "fix" a long-running issue involving flakes due to
creation/deletion of DCs with the same name in said tests by making sure
that the "old" DC won't accidentally interfere with the new DC.
Partial for #11004