Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm looking to monitor performance on my Linux servers (which happen to be Centos). What are the best tools for monitoring things in realtime such as:

  • Disk Performance I/O, swapping etc..
  • CPU Performance

Looking for low level tools, rather than web based tools such as Nagios, Ganglia etc...

n.b. I'd like to know exactly what each tool does rather than just having a list of random toolnames if possible please. Why the tool is a better option than others would be good also.

share|improve this question
up vote 3 down vote accepted

collectl is a great performance monitor with some unique and advanced features like InfiniBand and Lustre support or environmental and NFS monitoring. Of course it does basic stuff like disk and cpu monitoring, too. Highly recommended!

share|improve this answer
This is pretty cool, used it tonight on some disk I/O monitoring... working well... thanks for the suggestion... – Jon Oct 16 '09 at 1:54

iostat, sar, top, munin

share|improve this answer
Simple and to the point, I like it! – Dave Drager Oct 15 '09 at 18:01
+1 for munin, very easy to setup and use. – emills Oct 15 '09 at 21:16

I've recently release collectl-utils which provided graphing capabilities for collectl data with a web-based tool called colplot. The one key thing with this accuracy. Using gnuplot for the the graphics it's very fast and shows every data point, unlike RRD which 'normalizes' your data. -mark

share|improve this answer

Doing a Google search for "linux monitoring tools", you find some stuff.

One of the good links found: This article tells you what each tool does, and will give you an idea why one tool is better than another for a particular purpose.

The tools mezgani recommended are listed in that article, except Munin. Munin doesn't seem to be what you are looking for anyway (it is web-based and you specifically said you didn't want web-based).

share|improve this answer

If you want decent historical data graphing, especially for multiple systems, check out Cacti. It requires configuring SNMP on each system (easy).

share|improve this answer

If you want graph easy your Linux performance, try

Open source tool with web access and live graphs!

share|improve this answer
original question "Looking for low level tools, rather than web based tools" – Sirex Oct 7 '10 at 11:43

Use atop, that will give you percentage of how much each resource is busy. Also it color codes data and if something is botteling out system resources it will color code those entries.

In CentOS you can install it from EPEL/RPMForge repos.


share|improve this answer

I can see that you are looking for low level tools. But you might want to rethink as I believe performance monitoring is a regular exercise. From my experience I can tell how difficult it can become if you don't use dedicated tools. These are the tools I came across.

  1. Munin
  2. Zabbix
  3. SeaLion
share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.