Ran into something tonight I've not seen before. The system has been up
for 20 days, and has not missed a beat of any sorts, even when getting
beat on at a decent level. What my concern is, tonight I attempted to
log out so X would restart, and from the looks of all the stuff on the
screen (which never did reload gdm) was chock full of statements where
swap had been exhausted, and processes were being killed off. The only
way I was able to recover was to login via another machine and kill -9
the gdm process. Could not access any of the VT's at all. What I'm
wondering is if swap is not being released, rather if processes are not
being moved out of swap when the load decreased. From the looks of top
top - 21:20:59 up 19 days, 23:07, 5 users, load average: 0.22, 0.24,
0.17
Tasks: 146 total, 1 running, 145 sleeping, 0 stopped, 0 zombie
Cpu0 : 1.7% us, 0.3% sy, 0.0% ni, 98.0% id, 0.0% wa, 0.0% hi,
0.0% si
Cpu1 : 0.7% us, 0.0% sy, 0.0% ni, 99.3% id, 0.0% wa, 0.0% hi,
0.0% si
Cpu2 : 0.0% us, 0.0% sy, 0.0% ni, 100.0% id, 0.0% wa, 0.0% hi,
0.0% si
Cpu3 : 0.7% us, 0.0% sy, 0.0% ni, 99.3% id, 0.0% wa, 0.0% hi,
0.0% si
Mem: 2055916k total, 1199952k used, 855964k free, 50464k buffers
Swap: 1052216k total, 64020k used, 988196k free, 781740k cached
There seems to be plenty of swap free, even tho the cache is still OK.
I didn't have the presence of mind to run any other tools to see the
system behavior. Believe this is the first time I've had to kill gdm to
reload and recover. Perhaps there is some tuning or tweaking I could
do? The main processes are the wx model, which uses anywhere from 1.2
to 1.4G of memory, plus all the normal stuff you find on a workstation.
Suggestions anyone?
Thanks..
--
Sam Drinkard <sam at wa4phy.net>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.centos.org/pipermail/centos/attachments/20060210/9e492172/attachment-0002.html>
