up vote 1 down vote favorite
share [g+] share [fb]

The monitoring software we use has been flagged a box that serves two apps:

  • Confluence (xms: 1 GB, xmx: 2 GB)
  • JIRA (xms: 1 GB, xmx: 1 GB)

These are the only two applications on the box; nothing else but it is a VM.

Server Config

  • VM (VMWare)
  • RedHat EL 5
  • 2 GB memory allocated

The Graph

alt text What's really weird to me is that peak times should be between 8 AM and 5 PM, but on Thursday morning, from 1 AM to 8 AM. Not as bad, but Friday is similar--swap is being used though there's no load I'd expect at that time.

Help?

link|improve this question
Our servers fetch and apply updates via cron-apt during the wee hours (4AM). Anything like that going on? – nedm Oct 1 '10 at 21:55
@nedm: Not applicable to us, as far as I know. Last I heard, the VM hasn't been patched in months. I don't think it's been in use for 9 months yet. – OMG Ponies Oct 1 '10 at 21:57
feedback

1 Answer

up vote 3 down vote accepted

Run atop in daemon mode:

atop -a -w /var/log/atop.log 600

let it log for the night, read the log afterwards with

atop -r /var/log/atop.log

you can scroll forward and backward with t and T, memory details are shown after pressing m

link|improve this answer
Sorry for the delay. We determined that the applications were delegated more RAM than the system had, so it was increased and things are fine now--no more memory/swap warnings. – OMG Ponies Oct 10 '10 at 15:43
feedback

Your Answer

 
or
required, but never shown
discard

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