I have a requirement where in I need to monitor some custom services on aws nodes and collect metrics in timeseries. There are specifically two use cases. One being the monitoring of hardware resources like cpu, mem, disk util etc and the other being monitoring service specific metrics.
While reading up I came across collectd as one of the open source option. However I wanted to know how I can use collectd to monitor service specific metrics. Does collectd expose APIs which the service can use to log the metrics and if yes how performant is it.
I am new to collectd & would like to know if there are any other open source options as well.
The collectd agent can monitor custom metrics using read plugins: