-
Notifications
You must be signed in to change notification settings - Fork 148
Description
Describe the issue:
We introduced introduced a headless gateway service for the 8.8 chart, to distribute the load correctly.
Both services
- https://github.com/camunda/camunda-platform-helm/blob/main/charts/camunda-platform-8.8/templates/orchestration/gateway-service.yaml
- https://github.com/camunda/camunda-platform-helm/blob/main/charts/camunda-platform-8.8/templates/orchestration/service.yaml
Have the same labels, this causes issues when using a servicemonitor scraping for zeebe-broker. As this would duplicate the metrics. We need some better differentiation.
For example we recently went back to the zeebe-broker labels for some components, we might need to consider the same for the gateway service and set it to zeebe-gateway as it was for 8.7 see here https://github.com/camunda/camunda-platform-helm/blob/main/charts/camunda-platform-8.7/templates/zeebe-gateway/_helpers.tpl#L26
Actual behavior:
Using the select component:zeebe-broker causes to return both services
$ k get svc -l app.kubernetes.io/component=zeebe-broker
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
camunda ClusterIP None <none> 26502/TCP,26501/TCP,9600/TCP 25m
camunda-gateway ClusterIP 10.4.9.4 <none> 8080/TCP,9600/TCP,26500/TCP 25m
Expected behavior:
The services have labels to distinguish each other
How to reproduce:
Use the latest alpha8 chart version
Logs:
Environment:
Please note: Without the following info, it's hard to resolve the issue and probably it will be closed.
- Platform:
- Helm CLI version:
- Chart version: alpha8
- Values file:
Metadata
Metadata
Assignees
Labels
Type
Projects
Status