ok. for the past two months, i've been noticing that, with about 6 days of uptime on my server, my ram (512mb) would be low and lots of swap in use. i'm thinking memory leak in some code somewhere (only 3 sites on the raq). so, i'd just restart the raq. which i know is not the solution, but, i've been in the middle of a move, yadda yadda, blah blah....


sooo, just now, i ssh'd into the raq, and ran a top and got this:

55 processes: 37 sleeping, 18 running, 0 zombie, 0 stopped
CPU states: 72.9% user, 27.0% system, 0.0% nice, 0.0% idle
Mem: 517368K av, 398220K used, 119148K free, 177504K shrd, 186808K buff
Swap: 131532K av, 0K used, 131532K free 67976K cached

SO, i then login to my RAQ CP and do system status and get this:

Total memory on the system is 648900 KB
Free available memory is 479996 KB
Used memory is 168904 KB which is 26 percent of the total memory


do wha?

Trust Top.........

and do a ps- aux.


I get strange results with Raq CP also with one of the servers.
Like right now, Ive got a a server stating in the Raq cp that there 95% of the ram is taken.

So what do I do? go in and take a look at top and ps -aux.

Look at processes that have been running for extraordinary times and kill the PID.

What helps is just restarting httpd. At least it will restart and terminate runaway cgi scripts and the mem will go down a bit.

I believe the last resort is to reboot the server.

The Raq CP System Status adds together the physical and swap memory.
I suppose that the free available memory is the sum of Top free and shrd mem + free swap.