What needs to be configured on a Linux server in order to collect statistics on which applications are eating resources?

Question

linux

There is a Linux server, periodically it slows down for one reason or another. I would like to run automated monitoring on it, which collects statistics on which applications / commands eat a lot of processor / memory / disk access / traffic. How to do it?

Answer 1

htop iotop mytop

Answer 2

top > file
pick up via ssh or put it on the ball with a script
if you want more abruptly and with a reserve for the future - snmp monitoring

Answer 3

Install collectd + influxdb + grafana - for example, follow this tutorial - https://howitmake.ru/blog/ubuntu/168.html
Well, the network is full of data on how to do this. This is how we monitor the cluster.
Yes, there is also a nice netdata project - https://my-netdata.io
We look at it, but it cannot centrally store data (I can be wrong), but for one server it will certainly work for you. Works very fast, eats very little.

Answer 4

For a single netdata server, this is a great solution, give it a try.