CODEWITHSUNDEEP
grafanaprometheusscylla
soundararajan.c
soundararajan.c

Reputation: 2567

Monitor Scylla db using Prometheus and Grafana

I have installed scylla db in Server A.

Installed Prometheus and Grafana in Server B.

Monitoring the scylla db server resource(CPU , RAM , Disk perfromance ..,) using node exporter ( installed node exporter in Server A).

Need to monitor scylla db operations using Prometheus and Grafana(from Server B).

What are all the exporters needs to be installed in Server A , and dashboard needed for scylla db monitoring in Grafana

Upvotes: 3

Views: 1870

Answers (1)

Glauber Costa
Glauber Costa

Reputation: 676

There are two kinds of metrics relevant for the ScyllaDB monitoring stack:

a) Internal DB metrics. ScyllaDB exports prometheus metrics natively. So you don't need anything extra running in Server A for those to work. All you need to do is point to the server(s) in the scylla_servers.yml file from the monitoring side.

b) Linux metrics (disk utilization, network, etc). That is what node_exporter is for. So just make sure node_exporter is running in ServerA, and point to that server again in the node_exporter_servers.yml in the monitoring side.

In a lot of situations, of course, those two will be identical - only the ports differing (the exception is containerized environments where there are multiple ScyllaDB servers in a single Linux hosts). To fast track that case, Scylla Monitoring ships a script genconfig.py, which will generate node_exporter_servers.yml and scylla_server.yml with the correct ports given just a list of IPs of the servers running Scylla.

Upvotes: 8

Related Questions