Workload in holoinsight-server should use persistent storage volume

[ideal]

Describe this problem

In holoinsight-server, there are ceresdb, es, mongo, mysql, prometheus, currently none declares using persistent volume claim, so it fallbacks to anonymous docker volume specified in Dockerfile.

But if you delete the pod to restart it, or the pod is shifted to other node, the data will be lost, e.g. for holoinsight-mysql-0.

To Reproduce

kubectl delete po -n holoinsight-server holoinsight-mysql-0

Expected behavior

mysql data should be retained, but if you login into mysql, the databases and tables disappear.

Additional Information

No.

[xzchaoo]

Hi @ideal . Thanks for the feedback. As you said. The Chart currently provided is not enough for production-level deployment. Some improvements are needed in the future.