Skip to content

ci: bump enterprise deploy timeout to 15m (openshift flake investigation)#619

Closed
svietry wants to merge 1 commit into
mainfrom
sv/openshift-ci-timeout
Closed

ci: bump enterprise deploy timeout to 15m (openshift flake investigation)#619
svietry wants to merge 1 commit into
mainfrom
sv/openshift-ci-timeout

Conversation

@svietry

@svietry svietry commented Jul 10, 2026

Copy link
Copy Markdown
Contributor

Purpose (draft / investigation)

The Compatibility Matrix test (openshift, 4.18.0-okd) job has been failing on every PR that touches an enterprise-related chart (#617 k8s-inventory, #618 admission controller), while gke/aks and all kind tests pass. The failure is always in the Deploy enterprise step deploying the published anchore/enterprise chart, hitting helm's default 5-minute --wait deadline:

  • Error: INSTALLATION FAILED: timed out waiting for the condition
  • unexpected error when reading response body ... context deadline exceeded

This is unrelated to the chart bumps in those PRs — it's the enterprise stack not reaching Ready within 5m on OKD.

Change

  • Add --timeout 15m to the enterprise helm install --wait (both openshift + non-openshift branches) to rule out slow pod startup on OKD.
  • On deploy failure, dump kubectl get pods -o wide, describe pods, and namespace events so the actual stuck pod is visible (today the step dies before any diagnostics run).

This branch is based on the k8s-inventory bump so the workflow's path/enterprise-required gates actually trigger the openshift job.

Outcome to watch

  • Passes at 15m → root cause is just slow startup on OKD; we can keep a higher timeout.
  • Still fails → diagnostics will show which pod is stuck (SCC/image-pull/probe), pointing at the real enterprise-on-openshift regression.

🤖 Generated with Claude Code

@svietry svietry force-pushed the sv/openshift-ci-timeout branch 3 times, most recently from 7ecb261 to c7bc7d3 Compare July 10, 2026 20:37
The openshift compatibility-matrix job times out deploying enterprise at helm's
default 5m --wait; the stack needs ~6m to become ready on OKD. Bump --timeout to 15m.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: Shari Vietry <shari.vietry@gmail.com>
@svietry svietry force-pushed the sv/openshift-ci-timeout branch from c7bc7d3 to d506a50 Compare July 10, 2026 20:40
@svietry svietry marked this pull request as ready for review July 10, 2026 20:42
@svietry svietry requested a review from a team as a code owner July 10, 2026 20:42
Copilot AI review requested due to automatic review settings July 10, 2026 20:42

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR updates the OpenShift CI workflow to reduce flakes in the “Deploy enterprise” step by extending Helm’s wait timeout when installing the published anchore/enterprise chart, as part of investigating slow readiness on OKD/OpenShift.

Changes:

  • Increase Helm install wait timeout for enterprise deploys to 15 minutes (OpenShift and non-OpenShift branches).
  • Update the inline version comments for the Replicated cluster create/remove actions to v1.16.2 (commit pin unchanged).

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment on lines +140 to 144
helm install enterprise anchore/enterprise --namespace anchore -f stable/anchore-admission-controller/ci/enterprise-openshift-vals.yaml --wait --timeout 15m
else
echo "Installing enterprise"
helm install enterprise anchore/enterprise --namespace anchore -f stable/anchore-admission-controller/ci/enterprise-vals.yaml --wait
helm install enterprise anchore/enterprise --namespace anchore -f stable/anchore-admission-controller/ci/enterprise-vals.yaml --wait --timeout 15m
fi
@svietry svietry closed this Jul 10, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants