avirajkhare00 commented on issue #3007:
URL: https://github.com/apache/iggy/issues/3007#issuecomment-4115143345

   Follow-up on the implementation in #3019: I updated the runtime-validation 
approach from the earlier port-forward smoke test to an ingress-based smoke 
test.
   
   Current approach:
   - keep `helm/validate` for the fast render-level coverage: pinned Helm 
install, `helm lint --strict`, the 6 requested `helm template` scenarios, 
deterministic legacy API assertions, and explicit chart/image version assertions
   - add a separate `helm/smoke` task for runtime coverage
   - `helm/smoke` now creates a `kind` cluster, installs pinned 
`ingress-nginx`, installs the chart with both `server.ingress` and `ui.ingress` 
enabled, and probes the two components through ingress hosts instead of using 
`kubectl port-forward`
   - the smoke checks are `GET /ping` for the server ingress path and `GET 
/healthz` for the UI ingress path
   - the UI image tag is pinned from `web/package.json` during the smoke 
install so the job is not coupled to the rolling `:edge` tag
   - on failure, the workflow collects namespace resources, ingresses, 
controller logs, pod descriptions, and kind logs
   
   Reason for the change: validating through ingress gives real coverage for 
the chart path we actually expose and exercises both components together, 
instead of only proving that the backing services were reachable with local 
port-forwards.
   
   The latest upstream pre-merge run for this approach passed: `23471288322`.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to