Imported Cassandra detail .iam not able to see the metric node_cpu in my grafana

1

Hey,

Iam trying to implement Cassandra detail dashboard in my environment. when i try to setup the node_cpu metric iam not able to find the metrics. what c=kinda of metrics need to be added in prometheus .yaml file to show the node_cpu?
iam trying to do like the one “sum(irate(node_cpu{alias=”$instance"}[5m])) by (mode) * 100 / count_scalar(node_cpu{mode=“user”, alias="$instance"})" what need to be done to get node_cpu,node_memory_Buffers,node_memory_MemFree and node_load5 for say how would be able to pull all the node* metrics and also what needs to be done to get the alias as instance . sum(irate(cassandra_columnfamily_writelatency{alias="$instance"}[5m]))

2

Hi,

What Prometheus exporter are you using on your Cassandra host(s)?

Marcus

3

hi Marcus,

I’m using jmx_prometheus_javaagent-0.1.0.jar

Rahul

4

Hi,

Have you seen this community built dashboard? If I remeber correctly that is built based on the cassandra jvm exporter.

Please try it out and let me no if that works

Marcus

5

Hey,

I can export Cassandra stats but not node cpu stats

6

Maybe you’ll need to use the node exporter as well for that?

Marcus

7

Can we add node exporter in cassandra-env file?

8

Hi,

I don’t have that much Cassandra knowledge. When I tried Cassandra with Prometheus I built a custom docker container with cassandra and jmx exporter. For the host part in the docker scenario you probably want to monitor the container instead with cadvisor for example.

To give you some kind of answer to your question you probably run the node exporter side by side with cassandra and the jmx exporter if you run cassandra on a virtual or physical host.

Marcus