Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

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 have a virtual ubuntu ftp server that is sometimes extremely slow. Connecting via SSH takes forever, it is very slow to run commands, etc. Other times, it runs normally.

How can I troubleshoot the problem? I am 99% positive that it is the actual ubuntu installation, not the virtual machine, as we have many other virtual linux boxes that are lightning fast.

Also, any general ubuntu server speed-up links or suggestions are greatly appreciated.

share|improve this question
Did you check the resources usage? You did not provide us with enough useful information. – Khaled Sep 14 '11 at 13:47
What Virtualization system are you running? – uSlackr Sep 14 '11 at 16:02
VMWare 4.1 (ESXi) – divided Sep 16 '11 at 13:25

With a virtual guest, you have to look at the underlying server to ensure it is not over-allocated. I would look at the CPU, RAM and IO stats for the host to ensure your guest isn't starved for resources.

Is it possible some other guest is grabbing the CPU at certain time?

share|improve this answer
I've made sure that there are enough resources allocated to the VM. – divided Sep 14 '11 at 15:05

Does the output of 'top' show nearly 100% iowait by any chance? I have an Ubuntu server that was painfully slow under VMWare until I installed VMWare Tools, once I'd rebooted it after that it was absolutely fine.

share|improve this answer
No, I don't see 100% under iowait. This machine already has VMWare tools installed. – divided Sep 14 '11 at 15:05

A few things come to mind:

  1. slow DNS can slow downlogins as it tries to do a reverse lookup in your IP for the logs
  2. you need to keep an eye on the CPU load, the RAM load, whether or not SWAP is being heavily used, and how much IO there is.
  3. a nice rough-metric is the system load, a machine that's running at capacity will have a load of 1xnumCPUs, so an 8CPU system with a load of 4 is running at about half capacity. Servers can go over a capacity of 1-per-cpu, but that means processes are getting queued, so the system will be slow and laggy.

Generally what you want is data, as much of it as you can get. Don't suppose you have any sort of SNMP monitoring tools you could point at the VM? Like Cacti or ZenOS?

share|improve this answer
No, I don't have any of those monitoring tools. – divided Sep 14 '11 at 15:09
Then you'll have to manually look at the load etc when the server is being slow using the top command. You can test the DNS using the host or nslookup commands. – Bart B Sep 14 '11 at 16:18

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.