[rrd-users] Re: mrtg performance

Logg, Connie A. cal at SLAC.Stanford.EDU
Wed Sep 25 21:49:21 MEST 2002


You need to look for another way to do it.   MRTG does not scale well. One idea is to write scripts to collect the snmp stats and log them to flat files.  This collection can be done in less than 5 minutes.  Write an analysis script and plotting script which run asynchronously to the data collection. 

I have an older slower solaris system, and easily collect the data from 2600 switches in about 1.5-2. minutes.  You really do not need to plot the data every 5 minutes. In a crisis you can look at the raw data if you have it in a flat file, although there are other tools that can be used in a crisis situation.  Plotting the data every 20 minutes or so is usually frequent enough I find.  This will greatly reduce the load on your monitoring machine.  

Hopefully others can refer you to existing tools that scale better.

Connie Logg - Network Analyst - 650-926-2879
Stanford Linear Accelerator Center
MS 97; 2575 SandHill Road; Menlo Park CA 94025
"Happiness is found along the way, not at the end of the road"

-----Original Message-----
From: Augusto Castelan Carlson [mailto:carlson at npd.ufsc.br]
Sent: Wednesday, September 25, 2002 10:50 AM
To: 'rrd-users at list.ee.ethz.ch'
Subject: [rrd-users] mrtg performance



Hi!  I have i friend that sent me an email asking about a problem that he is having with mrtg. I´m going to translate with my poor english.
Perhaps something that may help him is to do the RDtool integration?

The email is:
####
I´m having some problems to monitor more than 4000 interfaces in habil time, or either, less than 5 minutes.  Analysing, mrtg takes most part of the time creating the images and generating the log archives.

I´m running MRTG in a XEON 1GHz single, with two RAID SCSI disks, 1GB of RAM (It´s a Compaq hardware) and running Windows 2000 server.

The total time to collect and generate the graphs of 24 interfaces of one switch gives a time 10 seconds, giving a total of approximately of 27 minutes.

What can I do to optimize MRTG to monitor this  volume of interfaces in a time less then 5 minutes? Sugestions?

Thank you!

Guilherme

########


Thank you!

Augusto Castelan Carlson
Florianópolis - SC - Brasil


--
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-users-request at list.ee.ethz.ch?subject=help
Archive     http://www.ee.ethz.ch/~slist/rrd-users
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi

--
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-users-request at list.ee.ethz.ch?subject=help
Archive     http://www.ee.ethz.ch/~slist/rrd-users
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi



More information about the rrd-users mailing list