[BUG] 2.7.4 - 2.7.5- monitoring/grafana - embedded or grafana links are 404

[lerminou]

Rancher Server Setup

• Rancher version: 2.7.4 or 2.75-rc5
• monitoring: rancher-monitoring:102.0.0+up40.1.2
• Installation option : docker

Information about the Cluster

• Kubernetes version: 1.23
• Cluster Type (Local/Downstream): downstream RKE imported

  Any embedded grafana ou full grafana link url returns a 404 not found

  

  Because when clilnkg on the grafana endpoint, we are redirected to https://rancher.intra.company.com/k8s/clusters/c-m-txpct7zr/api/v1/namespaces/cattle-monitoring-system/services/http:rancher-monitoring-grafana:80/proxy/d/rancher-node-detail-1/rancher-node-detail?orgId=1

  notice the cluster name in the url

  If I click on the grafana home button:

  then the url changes to : https://rancher.intra.company.com/k8s/clusters//api/v1/namespaces/cattle-monitoring-system/services/http:rancher-monitoring-grafana:80/proxy/?orgId=1

  the cluster name disappears and grafana is working To Reproduce

  Result

  Expected Result

  grafana is working for embedded links or full web Screenshots

  Additional context

  Maybe a side effect of : https://github.com/rancher/rancher/issues/41562 https://github.com/rancher/dashboard/issues/8885

[lerminou]

works well with grafana 7.5.8 (and it's nginx configuration) with the same rancher stack

[Delta1977]

Somethin else here: K3S v1.25.11+k3s1 with Rancher 2.7.4: URL Metrics :https://RANCHER_URL/k8s/clusters/local/api/v1/namespaces/cattle-monitoring-system/services/http:rancher-monitoring-grafana:80/ with 404 and changes to https://RANCHER_URL/api/v1/namespaces/cattle-monitoring-system/services/http:rancher-monitoring-grafana:80/proxy/?orgId=1 which works.

There is a difference between api endpoints

[HectorB-2020]

I'm afraid I have to confirm that this issue is observed in Rancher 2.6.13 with Kubernetes 1.24.15. Interestingly there are other issues here about error 404 in Grafana:

• https://github.com/rancher/rancher/issues/41562
• https://github.com/rancher/rancher/issues/41036
• https://github.com/rancher/rancher/issues/32884

[erSitzt]

Same here in 2.7.5, but different behaviour with different clusters.

Recent cluster with older monitoring version works v1.24.4+rke2r1 with rancher-monitoring:100.1.0+up19.0.3

new cluster with older k8s version but newer monitoring does not work v1.21.5 +rke2r1 with rancher-monitoring:102.0.2+up40.1.2