Karl E. Jorgensen wrote:
> Next time you compile things, start a couple of sessions (=separate
> windows):
> - vmstat 5 - to keep track of free memory and swapping
> - top - sorted so the most memory hungry processes are on top
> - tail -f /var/log/syslog - to see when oom-killer fires up
> - a compile session
>
> and keep an eye on what happens in the other sessions when gcc fails
I went ahead and performed the above. The compile session did crash a few times
as usual, but the oom-killer never appeared. In fact, looking through the logs,
it hasn't appeared since July 15th (the logs I posted earlier). So, I'm assuming
it appeared when I ran "memtest" with a memory value larger than the available
physical memory.
Here's a snapshot of "top" during the compile (sorted by memory usage):
top - 18:13:23 up 2 days, 7 min, 7 users, load average: 1.47, 0.95, 0.48
Tasks: 157 total, 3 running, 154 sleeping, 0 stopped, 0 zombie
Cpu(s): 95.7%us, 4.0%sy, 0.0%ni, 0.0%id, 0.0%wa, 0.0%hi, 0.3%si, 0.0
Mem: 2076768k total, 1847488k used, 229280k free, 7636k buffers
Swap: 1951888k total, 64k used, 1951824k free, 1532160k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
2389 cathyep 18 0 147m 53m 22m S 0.0 2.7 0:16.60 iceape-bin
6452 mythtv 25 0 60860 50m 5688 R 31.6 2.5 0:00.95 cc1plus
2620 mysql 21 0 126m 29m 5064 S 0.0 1.5 0:40.69 mysqld
17591 mythtv 18 0 223m 28m 15m S 0.0 1.4 1:47.32 mythbackend
3084 mythtv 15 0 45148 20m 14m S 0.0 1.0 0:01.00 gnome-panel
3087 mythtv 18 0 88012 19m 14m S 0.0 1.0 0:00.54 nautilus
2353 cathyep 18 0 37552 18m 11m S 0.0 0.9 0:00.50 gnome-panel
2354 cathyep 18 0 78476 18m 13m S 0.0 0.9 0:00.66 nautilus
2375 cathyep 16 0 61924 13m 10m S 0.0 0.7 0:00.29 gweather-applet
2916 root 17 0 25760 11m 5792 S 0.0 0.6 0:02.30 Xorg
3124 mythtv 23 0 31388 11m 9144 S 0.0 0.6 0:00.19 gnome-terminal
2276 cathyep 15 0 28892 10m 8232 S 0.0 0.5 0:00.18 x-session-manag
3004 mythtv 17 0 28744 10m 8452 S 0.0 0.5 0:00.17 x-session-manag
2339 cathyep 18 0 29836 9m 8144 S 0.0 0.5 0:00.17 gnome-settings-
3069 mythtv 18 0 29888 9.9m 8048 S 0.0 0.5 0:00.17 gnome-settings-
17963 mythtv 15 0 40228 8860 7400 S 0.0 0.4 0:00.52 gnome-cups-icon
3081 mythtv 19 0 14792 8056 6856 S 0.0 0.4 0:00.25 metacity
2350 cathyep 18 0 13988 8032 6196 S 0.0 0.4 0:00.53 metacity
2986 www-data 15 0 22276 7588 3812 S 0.0 0.4 0:00.09 apache2
The used memory went up and down slightly, during the compile. And, actually, not
that much different from the idle system. I notice that when I first start my
system, the used memory is just below one Gig. But over time (a day or two), it
creeps up to around 1.8 Gig. The used swap starts out at zero, and I don't
remember it ever going above 64k. I don't see anything in "top" hogging the memory.
When I finished the compile there was actually more free memory than before the
compile. So something released some memory. Here's a few snips of vmstat:
procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu----
r b swpd free buff cache si so bi bo in cs us sy id wa
<before>
0 0 64 79160 5988 1735280 0 0 33 102 206 47 2 1 97 0
<during>
2 0 64 100864 6432 1690732 0 0 71 131 355 328 96 2 0 2
...
1 0 64 165836 6276 1567052 0 0 0 11 333 299 99 1 0 0
...
1 0 64 243132 6324 1568664 0 0 4 66 340 271 27 2 71 0
...
1 0 64 368568 7828 1441996 0 0 0 434 400 770 10 9 82 0
<after>
0 0 64 370636 9384 1441772 0 0 33 101 207 49 2 1 97 0
I'm almost to the point of blowing the system away and installing Etch. Anyone
with insight would be appreciated.
Thanks,
Mike
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
