Description of problem:

checked on 4.19.0-0.nightly-2025-09-19-111328/4.20.0-0.nightly-2025-09-19-231410/4.21.0-0.nightly-2025-09-19-012417, admin user goes to "Observe -> Dashboards", "API Performance" is the default dashboards, select etcd dashboard then select "Kubernetes / Compute Resources / Cluster" dashboard, No datapoints found for the graph. checked  the expr for "CPU Utilisation" on the board, it is cluster:node_cpu:ratio_rate5m{cluster="etcd"} now, which the {cluster="etcd"} is carried from the etcd dashboard, 

if we click other memu, example "Observe -> Alerting" first, then click "Observe -> Dashboards", select "Kubernetes / Compute Resources / Cluster" dashboard directly, will see the result for the graphs. the expr for "CPU Utilisation" is cluster:node_cpu:ratio_rate5m{cluster=""}, see recording: 
https://drive.google.com/file/d/1sQ3jsEyjTcKWgd34_4WrN-MTjQ2azDdA/view?usp=drive_link

4.18 does not have such issue, compared with 4.18.0-0.nightly-2025-09-21-213014

for 4.18.0-0.nightly-2025-09-21-213014, dashboards url are

API Performance: https://<console_url>/monitoring/dashboards/grafana-dashboard-apiserver-performance?period=5m&apiserver=kube-apiserver
etcd: https://<console_url>/monitoring/dashboards/etcd-dashboard?cluster=etcd
Kubernetes / Compute Resources / Cluster: https://<console_url>/monitoring/dashboards/dashboard-k8s-resources-cluster

4.19+

API Performance:https://<console_url>/monitoring/dashboards/grafana-dashboard-apiserver-performance?timeRange=1800000&period=5m&apiserver=kube-apiserver

etcd: https://<console_url>/monitoring/dashboards/etcd-dashboard?timeRange=1800000&period=5m&apiserver=kube-apiserver&cluster=etcd
Kubernetes / Compute Resources / Cluster: https://<console_url>/monitoring/dashboards/dashboard-k8s-resources-cluster?timeRange=1800000&period=5m&apiserver=kube-apiserver&cluster=etcd

we can find out the 4.19+ dashboards carry the parameters from the previous dashboard, then append its dashboard parameters, this could cause the No datapoints found for the graph

Version-Release number of selected component (if applicable):

4.19+

How reproducible:

always

Steps to Reproduce:

1. see the description

Actual results:

"Observe - Dashboards" dashboards carry the parameters from the previous dashboard caused no datapoints found for the graph 

Expected results:

show result for graph

Additional info: