I'm using kubernetes with helm. My github pipeline contains:
- name: Verify deployment
run: kubectl rollout status deployment/name-api --namespace my-prod --timeout=60s
Sometime this part fails with:
Readiness probe failed: Get “https://...:80/”: dial tcp ...:80: connect: connection refused
When I then re-run the Action, it often passes.
What should I do against this?
I've already tried adding the last line below in deployment.yaml
in the templates
folder, based on the answer at https://stackoverflow.com/a/51932875. But that didn't help.
spec:
template:
spec:
containers:
- name: {{ .Chart.Name }}
…
readinessProbe:
initialDelaySeconds: 5