grafana / k8s-monitoring-helm

Apache License 2.0
136 stars 58 forks source link

[Discussion] Synthetic monitoring checks #283

Open danielloader opened 7 months ago

danielloader commented 7 months ago

First off, thanks for this chart! It's great to have a turnkey solution to onboarding a kubernetes cluster into grafana cloud. I also don't know where else to pose this question as discussions are disabled, nor did it make sense in the synthetic testing repos.

One hole in my monitoring stack is the synthetic checking - it's offered by grafana cloud but I have no way of cleanly automating adding and removing checks when I add/remove ingress backed applications to my cluster.

This chart uses prometheus CRDs for monitoring and scraping, and clearly that's not going to cover this usecase so I am asking what would be the best solution to handling this problem?

As far as I see it there's a couple options:

The worst solution but the only one I can manage myself right now is a ArgoCD PostSync hooked job container with the grafana api key and a curl command to add the check - it's not impossible just shifting the maintenance overhead into every single application I deploy.

Any thoughts are welcome as I'd like to integrate more tightly with grafana cloud from a kubernetes deployment!

skl commented 7 months ago

Nice idea! I've asked Synthetic Monitoring team for their thoughts on this.

bentonam commented 3 months ago

Adding to the discussion, I understand the reasoning behind the request. Wanted to add that the Grafana Terraform Provider does support synthetic monitoring checks. https://registry.terraform.io/providers/grafana/grafana/latest/docs/resources/synthetic_monitoring_check