[mrtg] Re: Has anyone used a RAMdisk to speed up MRTG?
Jim Johnson
jim at htch.com
Thu Jul 8 20:40:12 MEST 1999
> Terje Krogdahl wrote:
> > "Jim Johnson" <jim at htch.com> writes:
> >
> > > Right now I'm monitoring 25 devices. My Pentium 166 running
> RH Linux 6.0
> > > takes 3:40 (mins:secs) for 3 different instances of MRTG
> (each polling a 3rd
> > > of my devices) to finish running. If I run all of the
> targets from just one
> > > config (and therefore one instance of MRTG) the run time only
> goes up to 4
> > > mins.
> >
> > It seems to me like something causes MRTG to have a looong wait here.
> > Have you checked the output from MRTG to see if there's a problem? Maybe
> > some targets hold MRTG up?
> >
> > I used to do 365 targets (multiple instances) on a Pentium 200 (on IDE
> > disks, mind you!) with few problems.
>
> MRTG is very I/O intensive when it's updating the logs. 365 targets on
> IDE drives isn't too bad at all. I guess the question we should be
> asking Mr. Johnson is 'How many Targets are you logging with what
> kinds of drives?' The results of
> /proc/ide/ideX/hdX/<model,cache,settings>
> can be useful here.
>
I should've listed the # of targets instead of devices in my original post.
I'm currently monitoring 211 targets on my IDE WD drive (and I can't afford
anything better).
The hard drive is a WDC AC11000H
cache = 128
and settings is:
name value min max mode
---- ----- --- --- ----
bios_cyl 1023 0 65535 rw
bios_head 32 0 255 rw
bios_sect 63 0 63 rw
breada_readahead 4 0 127 rw
bswap 0 0 1 r
file_readahead 124 0 2097151 rw
io_32bit 0 0 3 rw
keepsettings 0 0 1 rw
max_kb_per_request 64 1 127 rw
multcount 0 0 8 rw
nice1 1 0 1 rw
nowerr 0 0 1 rw
pio_mode write-only 0 255 w
slow 0 0 1 rw
unmaskirq 0 0 1 rw
using_dma 0 0 1 rw
--
* To unsubscribe from the mrtg mailing list, send a message with the
subject: unsubscribe to mrtg-request at list.ee.ethz.ch
* The mailing list archive is at http://www.ee.ethz.ch/~slist/mrtg
More information about the mrtg
mailing list