Etcd Monitoring

I’m trying to use PMM to monitor a PostgreSQL cluster and so far this looks like an excellent product!

The cluster is build using Patroni and etcd which has a mechanism to integrate with Prometheus and Grafana, described here.

Can I use PMM’s VictoriaMetrics and Grafana components to scrap the etcd cluster and display the information?

For etcd I need to use static_configs which should be supported by VictoriaMetrics but I’m not entirely sure which file(s) should I modify on PMM for this purpose.

Any help is greatly appreciated!

Thanks,

Emil

Since you’re using a prometheus compatible exporter, you can use one of two options for PMM:

External exporter or External Serverless exporter. The difference between them is just “is there a pmm client available on the system in question” (if yes use External, if no use External Serverless and run on any system that has pmm2-client installed)

The details can be found here and I’m guessing you’ll execute something like:
pmm-admin add external --service-name=etcd --listen-port=9090 --metrics-path=/metrics
there are more options you can pass to improve filtering and the like
pmm-admin add {external | external-serverless} --help
but this should get your started. Note that it’s unlikely that your metrics will show up on default dashboards since our default graphs may not be aware of the metric names but you can use the explore functionality of grafana to find the metrics and create graphs and charts as you like.

1 Like

Excellent, I will give it a try and report here the results.

Thanks a lot for your help!

Emil

1 Like