Consul-helm: Add Prometheus Service Monitor Operator support

Created on 11 Jun 2020 · 4Comments · Source: hashicorp/consul-helm

Add Prometheus operator ServiceMonitor support like:

  serviceMonitor:
    enabled: true

enhancement themtelemetry

Source

sergeyshaykhullin

👍8

All 4 comments

@ishustava Does this issue is in someone's roadmap?

I can create PR with ServiceMonitor

It will monitor service/consul-server:8500/v1/agent/metrics?format=prometheus
I don't get how to use client metrics, does it has separate one, or consul-server is returning all?
P.S. As i understand it is returning all metrics required, so we have to have one root ServiceMonitor just for it?

How it might look like

server:
  extraConfig: |
    {"telemetry": {"prometheus_retention_time": "120s"}}

client:
  extraConfig: |
    {"telemetry": {"prometheus_retention_time": "120s"}}

serviceMonitor:
  enabled: true
  interval: ...
  scrapeTimeout: ...
  namespace: ...

sergeyshaykhullin on 16 Oct 2020

👍1

Hi Sergey,

I don't get how to use client metrics.

Some metrics only need to be retrieved from the consul servers but there are other metrics that both clients and servers have, e.g. consul.api.http, that would likely also be useful to collect across all clients and servers.

With the upcoming Consul 1.9 release, we are adding the ability to view prometheus metrics in the Consul UI so we will be looking at deeper prometheus support in the helm chart. So if you have a serviceMonitor config that you recommend we'd love to see a PR.