[Systems] Apache offline

Aleksey Lim alsroot at activitycentral.org
Mon Oct 3 15:04:32 EDT 2011


On Mon, Oct 03, 2011 at 02:17:08PM -0400, Bernie Innocenti wrote:
> On Mon, 2011-10-03 at 09:40 +0000, Aleksey Lim wrote:
> > On Mon, Oct 03, 2011 at 09:12:05AM +0000, Aleksey Lim wrote:
> > > Hi all!
> > > 
> > > About 04:20, Oct 3, apache was in offline.
> > > There are also another bad peaks for that time,
> > > http://munin.sugarlabs.org/sugarlabs.org/sunjammer.sugarlabs.org/index.html
> > > Any ideas?
> > 
> > eg, the question might be what programm like to call alloc()
> > and doesnt like to call free()
> > 
> > http://munin.sugarlabs.org/static/dynazoom.html?plugin_name=sugarlabs.org%2Fsunjammer.sugarlabs.org%2Fmemory&start_iso8601=2011-10-03T05%3A50%3A42-0000&stop_iso8601=2011-10-03T08%3A46%3A12-0000&start_epoch=1317621042&stop_epoch=1317631572&lower_limit=&upper_limit=&size_x=800&size_y=400
> 
> 
> I don't havetime to investigate now, but on sunjammer we have atop
> running. You can use "atop -r" to go back in time in the logs and find
> the errant process.

I did, it seems to be apache, at 2011/10/03  04:12:16, there were
440 Apache process and 56.7G of RSIZE. That was also a peak of ASLO
visits. For now there is MaxClients=512, maybe we need to make it lower
or there is a away how to minimize memory per client?

-- 
Aleksey


More information about the Systems mailing list