[mrtg] Re: mrtg performance

[Tobias Oetiker]

Hi Perry,

the 'data writing' part is not parallelized since it is disk-bound,
if you fork at that point in time mrtg will only hit the disks even
harder (an less organized) which will cause the performance to drop
even further ...

one way of solving this is to get faster disks ...

cheers
tobi


Today Perry, Duane wrote:

> I am running a huge cfg file.  It polls 28,000 interfaces.  I noticed
> that the SNMP query/response part of the process goes quickly.  I set
> forks=16 and you can see the 16 processes running simultaneously.  When
> it writes to the rrd files it drops back to a single process and takes a
> long time.  I write tiny klunky perl scripts but have little experience
> and no training.  Is there some way to fork that part of the script?
> MRTG would be much more scaleable if that piece were more efficient.
>
>
> Duane Perry
>
> University of Missouri
>
>
> --
> Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
> Archive     http://lists.ee.ethz.ch/mrtg
> FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
> WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi
>
>

-- 
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten
http://tobi.oetiker.ch tobi at oetiker.ch ++41 62 213 9902

--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://lists.ee.ethz.ch/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi