Import master to stable-1.5 #509
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* Move the SNMP trap delivery checks Move the SNMP trap delivery checks as where they are situated now seems to cause false positives. Moves the checks closer to the end of the smoketest run seems to result in a better change that the logs the check is looking for have been provided. * Use a loop to check for SNMP status with break and max time
This reverts commit af35714.
* Make all certs 8yr expiry * Use certificate_duration and test against generated cert * Better messages during CI cloning
* Expand support for OCP 4.11 Allow installation to be done on OCP 4.11 while updating the smoketest jobs to support later versions of the client. Also migrate to using community-operators CatalogSource instead of OperatorHub.io. Only enable community-operators when the use_community strategy is enabled. Update the token request syntax when requesting a service account token. Add checks to look for oc client version and fail if we're using a version that's too old. * Make passwords safer in smoketest job template Encapsulate the password values with double quotes to help make them safer for consumption in the template. I had an odd situation where the password contained a bunch of extended characters and caused the smoketest to report an error on the template having an issue with yaml to json. The password contained several characters such as . and : which confused the template. Wrapping the contents in the double quotes allowed the smoketest to apply the job.batch template and result in a working smoketest run.
* Replacing the placeholder namespace during the build results in a "there are local changes" error on next build * This forces the checkout to discard that (and other!?) local changes * Quicker dev/test loop
* Update oc to 4.11 in jenkins agent Need 4.11 for new token handling changes
Remove the OperatorHub.io CatalogSource and instead use the community-operators CatalogSource which is available with an OCP installation. Ideally this will avoid some of the conflicts we've been seeing in our CI environment. This is a short term fix as future development will likely make use of Observability Operator to provide the metrics data store and alert delivery mechanism.
* Catalog changes * CI change to pre-clean cert-manager-operator * not 100% sure this is 4.12 related, but it's new and first seen during testing 4.12
* Remove Loki from stf-run-ci * Return "Get new operator sdk" to stf-run-ci
The GitHub Actions checkout action v2 is deprecated and needs to move to version 3.
* Implement SNMPtrap delivery controls Implement ability to override the default values for the SNMPtrap alertmanager receiver via prometheus-webhook-snmp component. Closes: STF-559 * Run operator-sdk generate bundle Run the following command to update the bundle artifacts: operator-sdk-0.19.4 generate bundle --metadata --manifests --channels unstable --default-channel unstable * Build out the remaining SNMP options Build out the remaining options for prometheus-webhook-snmp to allow for finer grained controls and delivery of SNMP traps via alertmanager alerts. * Generate bundle contents with operator-sdk
* Implement changes for operator-sdk-1.26.0 testing Implement changes that allow testing validation via operator-sdk-1.26.0 without bumping the entire bundle generation process from operator-sdk-0.19.4 to post-operator-sdk-1.x. These are the same tests run for validation during product pipeline verification. * Adds test to verify building of the bundle image works. * Adds KinD deployment to allow executing scorecard checks. Related: STF-1252 * Fix properties.yaml * Simplify use of RELEASE_VERSION variable (#412) * Add note about why we're copying files in
* Adds duration param for CA and endpoint certs Replaces certificate_duration for ca_certificate_duration and endpoint_certificate_duration. Set default value for those to 70080h (previous value) Removes the certificate_duration param from the Issuer resource since it's not actually needed (see [0]) [0] https://cert-manager.io/docs/reference/api-docs/#cert-manager.io/v1.IssuerConfig * Exposes CA and endpoint certificate duration config Exposes certificate duration config for both ElasticSearch and QDR Keeps the default value in use for now. Better default values should be discussed to be included in a follow up change. * Fix identation for certs duration param in servicetelemetry crd * Adds cert duration to the OLM catalog Includes cert duration params in the OLM catalog for both ElasticSearch and QDR * Changes snake_case to camelCase to yaml case Fix to match style convention * Adds pattern expresion for certs duration * Add certificates param to events and transport * Exposes duration parameter in the CI script Adds the duration parameter for both ElasticSearch and QDR in the CI script Also updates the OLM Catalog with the latest changes (certificates object) * Corrects naming to certificates params in CI script * Fix snake cae in the CI script params for cert duration * Fix identation for transports in the deploy_stf CI script --------- Co-authored-by: Chris Sibbitt <[email protected]>
* Fix PROMETHEUS_K8S_TOKEN to account for oc version mismatch * Fix to use correct variable name in test
* Move to newest oauth-proxy container * Move to bcrypt for htpasswd * Up/Downstream image handling for new oauth-proxy container * Skip broken ansible lint for htpasswd * Hack to add EPEL in upstream builds
…ion (#402) * [jenkins] Add custom context labels for github build status notification Multiple jenkins deployment can now be run and report their build status separately instead of both reporting to the same ``continuous-integration/jenkins/pr-merge`` and overriding each other. There is now ``continuous-integration/jenkins/ocp-<OCP_VERSION>/pr-merge`` NOTE: the OCP_VERSION is hardcoded at the moment https://github.com/jenkinsci/github-scm-trait-notification-context-plugin * [jenkins] Make the build status label configurable Added the OCP_VERSION var to the casc-configmap.yaml, so that the correct label can be set for the jobs --------- Co-authored-by: Chris Sibbitt <[email protected]>
Default to Observability Operator for Prometheus & Alertmanager
- Adds two options to observability_strategy:
* use_redhat (OBO Only)
* use_hybrid (OBO + friends)
- Default to redhat supported components in deployment
- Use upstream source for ObO in CI
- Added sensubility to deployment validation
- Added required RBAC for OBO usage
-When no explicit observability_strategy is set:
* Existing STF objects get an explicit "use_community" added
* New STF objects get the default ("use_redhat") explicitly added
- Narrow the scope of the smoke test (#422)
Co-authored-by: Leif Madsen <[email protected]>
* Use stable-v1 cert-manager in CI for OCP >= 4.12
* Currently we are installing :latest * We and OBO currently install v2.43.0 * This will mitigate any potential for a prometheus roll-back * We will remove the pin in STF 1.5.5 after migration to OBO is complete
* The check for a missing observabilityStrategy was faulty * STF object would be updated to `observabilityStrategy: use_community` any time there was a community Prometheus deployed
Allow bundle deployments to happen from unauthorized container repositories.
* [stf-run-ci] Generate extra logs if preflight checks fail * Update preflight checks
Bumps [oauthlib](https://github.com/oauthlib/oauthlib) from 3.2.0 to 3.2.2. - [Release notes](https://github.com/oauthlib/oauthlib/releases) - [Changelog](https://github.com/oauthlib/oauthlib/blob/master/CHANGELOG.rst) - [Commits](oauthlib/oauthlib@v3.2.0...v3.2.2) --- updated-dependencies: - dependency-name: oauthlib dependency-type: direct:production ... Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Emma Foley <[email protected]>
* [stf-run-ci][create_catalog] Swap query for a command task Using query looks up the kubeconfig on localhost, rather than the host that ansible is executing against. This behaviour is different from either using the shell/command modules or using k8s modules. For consistent behaviour, the queries are replaced with an alternative way to get the same information that will have consistent behahaviour whether executing against localhost or a remote host.
* Add requires infrastructure annotations Add required infrastructure annotations for the bundle. Implementation is done in generate_bundle.sh because the annotations.yaml file in the deploy/olm-catalog/ directory is not read by operator-sdk-0.19.4. Append required additional feature annotations to the generated annotations.yaml by operator-sdk generate bundle. Related STF-1530 * Include annotations in the CSV directly * Revert "Add requires infrastructure annotations" This reverts commit c9e9b2a. * Generate CSV contents with operator-sdk
* [stf-collect-logs] Add a role for log collection * Update build/run-ci.yaml * [stf-collect-logs] Update the resource name in * [stf-collect-logs] Update README * [stf-collect-logs] Remove unnecessary lines * ci/post-collect_logs: Use stf-collect-logs role * [stf-collect-logs]: Use namespace in oc commands --------- Co-authored-by: Chris Sibbitt <[email protected]>
* [zuul] Add job to deploy from nightly bundles This job doesn't build STF, but deploys from the pre-built and published bundles. This is useful to be able to do periodically to make sure our latest bundles are deploying, and no dependencies are out-of-date, for example
Support STF 1.5 from OCP 4.11 through 4.14 for the next release as OCP 4.10 is now EOL.
* Update base image to pass security scans Update the base image with a dnf update (need to excluse ansible because ansible updates aren't compatible with the current build). This keeps packages up to date to allow the resulting image to pass registry security scans at the expence of image size. * Clean up intermediate layer Co-authored-by: Chris Sibbitt <[email protected]> * Add comments to help understand Dockerfile readout * Spellcheck fix --------- Co-authored-by: Chris Sibbitt <[email protected]>
Add cluster observability operator as the preferred dependency (bottom of list is highest priority) when installing Service Telemetry Operator. The cluster-observability-operator is the name of the downstream (product) bundle in the Red Hat Operators CatalogSource. If installing for upstream, preferred operator will be observability-operator (when the Red Hat Operators CatalogSource is not available or enabled). And then as a fall-back method when neither Observability Operator or Cluster Observability Operator is not available, allow for Prometheus Operator from the Community Operators to satisfy for the Prometheus storage backend.
When deploying from the UI, don't populate the events SGs by default as they are no longer used in a default configuration in RHOSP.
I forgot to run a generate bundle yesterday in some furious patch work.
|
@elfiesmelfie not sure if we'll need changes here to get this to pass/land? |
|
Build failed (check pipeline). Post https://review.rdoproject.org/zuul/buildset/812315fa1a4d46128eaaaba40660e885 ❌ stf-crc-latest-nightly_bundles MERGE_CONFLICT in 2s |
Drop .zuul.yaml for stable-1.5 since it's not setup for non-main testing at this point. In the future we may develop a separate set of tests for the stable-1.5 branch during merge, but not for this initial import. We'll rely on Jenkins testing for our functional validations.
elfiesmelfie
approved these changes
Oct 27, 2023
|
The old "Linting" test was updated and renamed, so I'm skipping it (since it can't be run). |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Import changes for STF 1.5.3