lkml.org 
[lkml]   [2001]   [Jul]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: dead mem walking ;-)
On Thu, 12 Jul 2001, Marcelo Tosatti wrote:

> > > And that is what is eating the system performance.
> >
> > does it bring up the load up to 30 and make the machine unusable?
> > (kswapd was also sometimes in the top-list of CPU hogs, but since i
> > sorted it by memory...)
>
> Yes. Obviously that should not happen.
>
> What you're seeing _is_ a problem.

thx for the ACK. :)) in the beginning of the week i felt like
nobody really cares.

> > > Can you please show us the output of /proc/meminfo when the system is
> > > behaving badly ?
> >
> > hold on, we set s.th. up....
>
> Ok, thanks.

here (attachement) it is, first the "top" output (sorry for the
length, but i don't want to cut too much. even more info i can make av.
upon request).

interesting is, that after i killed one job, some minutes later the
the machine went to sleep again for a ~20 minutes:

4:49pm up 1:55, 2 users, load average: 21.05, 18.29, 16.79
55 processes: 52 sleeping, 3 running, 0 zombie, 0 stopped
CPU0 states: 0.1% user, 99.115% system, 0.22% nice, 0.59% idle
CPU1 states: 0.4% user, 99.106% system, 0.64% nice, 0.23% idle
Mem: 4057236K av, 4050648K used, 6588K free, 0K shrd, 2884K buff
Swap: 14337736K av, 2785436K used, 11552300K free 3428484K cached
PID PPID USER PRI SIZE SWAP RSS SHARE D STAT %CPU %MEM TIME COMMA
1374 1295 usersid 15 2140M 1.2G 924M 353M 0M R N 99.7 23.3 77:56 ceqsim
2089 760 root 19 1048 4 1044 824 55 R 0.2 0.0 0:00 top
1 0 root 8 76 12 64 64 4 S 0.0 0.0 0:02 init

a few secs later i did cat /proc/meminfo (machine was in the middle of
waking up again):

total: used: free: shared: buffers: cached:
Mem: 4154609664 4148756480 5853184 0 2953216 3529117696
Swap: 1796939776 2871406592 3220500480
MemTotal: 4057236 kB
MemFree: 5716 kB
MemShared: 0 kB
Buffers: 2884 kB
Cached: 3446404 kB
Active: 378132 kB
Inact_dirty: 2944608 kB
Inact_clean: 126548 kB
Inact_target: 560 kB
HighTotal: 3211200 kB
HighFree: 2036 kB
LowTotal: 846036 kB
LowFree: 3680 kB
SwapTotal: 14337736 kB
SwapFree: 11533628 kB


readprofile showed even worse statistics for __get_swap_page afterwards:

884565 total 1.0268
488027 __get_swap_page 897.1085
385109 default_idle 7405.9423
1920 do_anonymous_page 6.2338
1198 file_read_actor 4.8306
872 si_swapinfo 5.0698
782 do_softirq 5.7500
779 skb_copy_bits 0.9143
317 statm_pgd_range 0.8087
302 __rdtsc_delay 10.7857


why was the machine hung with that one job? memorywise that job was
only 2.1GB on my 4GB machine. is it also due to the __get_swap_page
issue or are the "Inact_dirty" pages related to that?


cheers,
~dirkw






[unhandled content-type:application/x-gtar]
\
 
 \ /
  Last update: 2005-03-22 12:57    [W:0.050 / U:0.156 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site