list events additionally to pod logs

test to find out reason of not-even-starting containers

.gitlab-ci.yml

@@ -34,8 +34,7 @@ integration-test:
     - k3d cluster create --volume $(pwd)/data:/k3d-host/ testgitlabci-${CI_JOB_ID} --agents 1 --wait
     - kubectl cluster-info
     - kubectl apply -f testing/pvc-redis.yaml
-    # default timeout is 5 min, move to 8m for some leeway
-    - helm install my-helm-release . --wait --timeout 20m --values ./testing/values-testing.yaml --values local-secrets.yaml
+    - helm install my-helm-release . --wait --timeout 10m --values ./testing/values-testing.yaml --values local-secrets.yaml
     - kubectl get po -A
   script:
     - cd testing && ./run_integration_test.sh
@@ -45,6 +44,8 @@ integration-test:
     # fetch all pod logs
     - mkdir "${CI_PROJECT_DIR}/logs"
     - kubectl get po -o name | while read POD ; do kubectl logs "$POD" >> "${CI_PROJECT_DIR}/logs/$(echo $POD | cut -d'/' -f2).txt" ; done || true
+    # fetch all pod events - to troubleshoot not-starting pods
+    - kubectl get po -o name | while read POD ; do kubectl get events --field-selector involvedObject.name="$POD" >> "${CI_PROJECT_DIR}/logs/$(echo $POD | cut -d'/' -f2).txt" ; done || true
     - k3d cluster delete testgitlabci-${CI_JOB_ID}
   artifacts:
     when: always