Repository navigation
Troubleshooting Horizontal Pod AutoScaling (HPA) Issues in Keycloak Deployment #37223
|
Hi all, I’m facing an issue with Horizontal Pod Autoscaling (HPA) in my Keycloak v26 setup and would appreciate some insight. I’m using Keycloak v26 with However, I’m encountering an issue where the instance value in the Keycloak Custom Resource (CR) is preventing the autoscaler from scaling the pods. A Keycloak pod is created, and then immediately terminated - similar to what is described in this SO question. It seems that the instant count in the Keycloak CR is conflicting with HPA functionality, blocking the scaling operation. I came across a related issue with the legacy operator where a property was introduced to disable replica synchronization between the Keycloak CR and the StatefulSet (see: keycloak/keycloak-operator#549). Is this issue still relevant to the current version of the Keycloak Operator? I wasn’t able to find any similar properties (or alternative solutions) in the latest documentation, specifically for Keycloak Operator v26.0.6, or in the YAML schema for the Keycloak CR. Here are my versions: Any guidance or suggestions on how to resolve this issue would be greatly appreciated! |
Replies: 1 comment 4 replies
|
Don't target the statefulset. The Keycloak resource is scalable. |
It should be able to use the built in metrics as the pods will be found via the label selector.
I don't believe one was ever added to the documentation.