The operator manages multiple TiDB clusters, and the QPS display in the dashboard is chaotic

[wxiaoke]

Bug Report

Please answer these questions before submitting your issue. Thanks!

What did you do?

reference:https://docs.pingcap.com/tidb-in-kubernetes/dev/monitor-a-tidb-cluster#monitor-multiple-clusters

1. I deploy 2 cluster with 1 operator: (cluster1, ns1; cluster2,ns2），and deploy monitor using the reference
2. Stress test on cluster1

What did you expect to see? Cluster1's grafana monitoring and dashboard have pressure. The grafana monitoring of cluster2 has no pressure, but the dashboard has the same pressure as cluster1.

What did you see instead? Cluster2's dashboard has no pressure

What version of TiDB Dashboard are you using (./tidb-dashboard --version)? 6.1.1

[wxiaoke]

maybe need to modify this.