0

I have a linux server box that runs the following stack:

  1. apache 2.2.9
  2. passenger 2.2.8
  3. rails 2.3.8
  4. MySQL 5.1

Recently we've encountered some machine performance issues, where the entire server became extremely slow (for web requests, db queries, file access, etc) to unresponsive at times, then returned to function normally.

We can tell the days and time-ranges when this happened, and we'd like to investigate further before running off and buying a new machine. The server runs on a small company intranet with no more than a couple of dozen requests per minute, and the hardware is reasonable (3 GHz core 2 duo processor, 8G RAM, 2 years old).

I was thinking to look into some logs around those dates to see if there was heavy traffic, machine load or excessive DB activity around those times. Any recommendations on how to investigate this issue and where to focus my efforts? thanks.

3 Answers 3

1

There is an article which we published online at: http://www.anchor.com.au/hosting/development/HuntingThePerformanceWumpus which could give you some hints.

Would recommend checking that out as a starting point.

2

I'd go for collectd and collectd-web as the WebGUI if you need one, Kcollectd if you don't need "realtime" capabilities.

1
  • I don't know, I find collectd-web unusable as of this time of writing. I prefer the bundled collection3 where it's so easy to select graphs you need.
    – syaz
    Aug 24, 2011 at 10:09
1

Add a monitoring system like Monit to know what happen on your server. There are a lot of monitoring capabilities.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .