PX-Central is waiting for Prometheus / Grafana and Portworx

If you are attempting to start PX-Central on a cluster without Prometheus and Grafana and/or Portworx
Currently, PX-Lighthouse will get into a crash loop. This is because Lighthouse currently expects all of these to be present in the cluster.

The workaround for Prometheus and Grafana is:
Change portworx/lh-config-sync:0.3 to portworx/lh-config-sync:0.4

The Portworx Service dependency fix is currently in review along with some other optimizations such as:

  • Prometheus CRD installation moved to lh-config-sync
  • Lighthouse Service creation moved to lh-config-sync
  • AWS detection mechanism
    • Lighthouse Service Type optimization
    • Prometheus and Grafana Service Type optimization
1 Like

Hi Paul,

I am running Rancher Kubernetes and have already installed Prometheus and Grafana, so I don’t select monitoring when creating my spec in PX Central. PX-Lighthouse monitoring is stuck on ‘Wait’ when I enter the URLs for monitoring. These URLs are setup automatically in Rancher and have Cluster IPs. Do you know what URLs I need to use for this to work?

These are the URLs Rancher has that generated:
Prometheus
https:///k8s/clusters/c-mj7bf/api/v1/namespaces/cattle-prometheus/services/access-prometheus:80/proxy/graph

Grafana
https:///k8s/clusters/c-mj7bf/api/v1/namespaces/cattle-prometheus/services/http:access-grafana:80/proxy/

Thanks.