Process swapping issue

[nyck]

I have a server that is running Red Hat Enterprise and Nagios was complaining that swap was 100% used. I verified this by running free -m.

How can I locate which process is causing this issue?

I tried running a top and nothing showed up!

 

[elgrandeperro]

are you using tmpfs? that takes away swap space.

 

[ciman]

Did you already try to limit the size of tmpfs in your /etc/fstab?

Does "vmstat 1" show any swap activity?

Is the output of "ps aux" similar to top?

 

[Annihilannic]

The output of sar -r can be very useful for diagnosing the source of memory usage, especially over time (if you are running sar data collection)

ps -eo pid,vsz,rsz,comm | sort -rn -k 2,2 | head should help identify the biggest virtual memory users on your system.

In top did you sort the processes by memory usage instead of CPU?

Annihilannic.

 

[Annihilannic]

I meant to type args in that ps command-line instead of comm... it'll make the processes easier to identify by printing the full commands, not just the process name itself.

Annihilannic.