-
Notifications
You must be signed in to change notification settings - Fork 1.9k
ci-operator/step-registry/manifests: Factor day-2 manifests into own step #8953
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.
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
ci-operator/step-registry/manifests: Factor day-2 manifests into own step #8953
Conversation
96d9a98
to
990de82
Compare
LGTM except we'll have to use files ( |
990de82
to
78df4fc
Compare
78df4fc
to
da26f4f
Compare
|
…step These have been part of the e2e test step since that step landed in a2fd8c3 (step-registry: add Origin E2E test step, 2020-01-31, openshift#6965). But pushing manifests is more of an install thing than a test thing. For example, if a user swaps in their own tests, we still want the cluster to push insights if their cluster profile includes the token). The logic I'm replacing was just handling the insights operator. To allow us to recycle the same logic for additional manifests, I've carved out generic handling for ${CLUSTER_PROFILE_DIR} and ${SHARED_DIR}. The day-2 naming avoids conflicts with the ipi/install step that is already consuming ${SHARED_DIR}/manifest_*.yml Glob expansion should have stable ordering; from POSIX [1]: If the pattern matches any existing filenames or pathnames, the pattern shall be replaced with those filenames and pathnames, sorted according to the collating sequence in effect in the current locale. Ideally we'd update the cluster profiles themselves to store the config in day-2-manifests-insights-live.yaml (via a core-services/ci-secret-bootstrap/_config.yaml edit?). But we can't do that until we've removed or ported the old-style templates under ci-operator/templates, so I'm leaving that off for now and keeping a HACK check to copy the current profile's file into the newly-expected target location. The previous logic also used an error-ignoring '|| true' since at least b887d20 (Deploy the insights secret for the support-operator in all templates, 2019-06-04, openshift#3951). That commit didn't motivate its use, so I'm dropping it here. If the step fails, figuring out the culprit should be pretty straightforward. [1]: https://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#tag_18_13_03
da26f4f
to
67cb826
Compare
@wking: The following tests failed, say
Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR. 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. I understand the commands that are listed here. |
/lgtm |
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: bbguimaraes, wking The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
@wking: Updated the following 2 configmaps:
In 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. |
These have been part of the e2e test step since that step landed in a2fd8c3 (#6965). But pushing manifests is more of an install thing than a test thing. For example, if a user swaps in their own tests, we still want the cluster to push insights if their cluster profile includes the token).
The logic I'm replacing was just handling the insights operator. To allow us to recycle the same logic for additional manifests, I've carved out generic handling for
${CLUSTER_PROFILE_DIR}/manifests
and${SHARED_DIR}/manifests
. There's a bit of hoop-jumping to buildDIRECTORIES
, becauseoc apply -r -f ...
chokes on empty directories:Ideally we'd update the cluster profiles themselves to store the config in
manifests/insights-live.yaml
(via acore-services/ci-secret-bootstrap/_config.yaml
edit?). But we can't do that until we've removed or ported the old-style templates underci-operator/templates
, so I'm leaving that off for now and keeping aHACK
check to copy the current profile's file into the newly-expected target location.The previous logic also used an error-ignoring
|| true
since at least b887d20 (#3951). That commit didn't motivate its use, so I'm dropping it here. If the step fails, figuring out the culprit should be pretty straightforward.