Re: Swap is eaten up
From: Kasper Dupont (kasperd_at_daimi.au.dk)
Date: 07/19/05
- Next message: Kasper Dupont: "Re: Interprocess communication"
- Previous message: David Schwartz: "Re: Swap is eaten up"
- In reply to: Uwe Goellner: "Swap is eaten up"
- Next in thread: Uwe Goellner: "Re: Swap is eaten up"
- Reply: Uwe Goellner: "Re: Swap is eaten up"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Tue, 19 Jul 2005 15:08:12 +0200
It is a bit unclear what your problem is. If the
problem just is, that you are running out of swap,
then the solution would be to add more swap.
If there is a leak, then that will of course not
help. In that case take a look on the size of the
virtual memory of the different processes:
grep VmSize /proc/*/status
And when you have a candidate, which may be the
culprit, then take a look on the maps pseudofile.
If the system is slowed down because of constant
disk access, then look for processes in non-
interruptible sleep.
If memory presure is too high, then add more
memory or look for the processes with the highest
residen set sizes.
-- Kasper Dupont -- der bruger for meget tid på usenet. Note to self: Don't try to allocate 256000 pages with GFP_KERNEL on x86.
- Next message: Kasper Dupont: "Re: Interprocess communication"
- Previous message: David Schwartz: "Re: Swap is eaten up"
- In reply to: Uwe Goellner: "Swap is eaten up"
- Next in thread: Uwe Goellner: "Re: Swap is eaten up"
- Reply: Uwe Goellner: "Re: Swap is eaten up"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|