On Tue, 28 Nov 2000, Peter Gervai wrote:
> Hello,
>
> I noticed the following:
>
> # config file:
> mem_max 16m
> lo_mark 8m
>
> # ps awux:
> USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
> proxy 3314 0.1 22.9 62780 29408 ? S Nov22 8:24 /usr/sbin/oops -c
>/etc/oops/oops.cfg
>
> I got used to the feeling when squid ("limited" to 8MB in-memory
> objects) started to use 80MB+ or more memory. Oops so far was
> slight on memory but had only light use. Now it's under fair amount
> of use, and started to use memory, growing slower than squid used to
> but still noticeable after a time.
>
> How shall I interpret the mem-max and lo-mark settings? 67MB virtual
> allocation size is obviously a bit higher than 16MB in the setting,
> or the 8MB at lo_mark. :) [even resident size is 29M, which makes
> sometimes to the swap, which isn't really good for a busy cache.]
If you are under silaris, I can advize to use ptmalloc. If you are under
Linux, ptmalloc is already there. Please, remember that lo_mark count only
document volumes, it doesn't count for any other structures, database page
cache.
VSZ sometimes can not move down even if you free almost all allocated
memory. ptmalloc handle this situation better then some other mallocs, it
can return unused memory to system. The most important thing is that vsz
must stop at some moment.
>
> (Is there a point to make a stat about memory allocation?)
I just don't know how to do this in portable way.
>
> cya,
> grin
> =====================================================================
> If you would like to unsubscribe from this list send message to
> [EMAIL PROTECTED] with "unsubscribe oops-eng" in message body.
> Archive is accessible on http://www.paco.net/oops/
>
Igor Khasilev |
PACO Links, igor at paco dot net |
=====================================================================
If you would like to unsubscribe from this list send message to
[EMAIL PROTECTED] with "unsubscribe oops-eng" in message body.
Archive is accessible on http://www.paco.net/oops/
