[mrtg] MRTG and RAM Disk

[Steve Shipway]

Have you considered using rrdcached?  This gives a similar improvement in response times.  In your case, where you are not so interested in the history, only Nagios checking the latest value, you can set the rrdcached time to be fairly large and effectively keep everything in memory...

Steve

Steve Shipway
University of Auckland ITS
UNIX Systems Design Lead
s.shipway at auckland.ac.nz<mailto:s.shipway at auckland.ac.nz>
Ph: +64 9 373 7599 ext 86487

________________________________
From: mrtg [mrtg-bounces+s.shipway=auckland.ac.nz at lists.oetiker.ch] on behalf of Troy Lea [troy_lea at hotmail.com]
Sent: Friday, 30 January 2015 6:13 p.m.
To: mrtg at lists.oetiker.ch
Subject: [mrtg] MRTG and RAM Disk

I've been playing around with the idea of having MRTG use a RAM Disk.

Specifically on Nagios XI and CentOS.

I've been successfully able to define the current settings and it does work well.

LogDir: /var/nagiosramdisk/mrtg
ThreshDir: /var/nagiosramdisk/mrtg
WorkDir: /var/nagiosramdisk/mrtg

Using the RAM Disk improves performance, especially once you get into 4000+ ports being queried.

The only problem I have is when I reboot the server the port files don't exist until the MRTG has run the first time. For example these files:

ub04_93.rrd
ub04_94.rrd
ub04_95.rrd
ub04_96.rrd
ub04_97.rrd
ub04_98.rrd

This causes a problem with the check_rrdtraf program that looks for these files. The problem is overcome once MRTG runs the first time.

I'm wondering if MRTG has an option which will generate all these files without actually doing an SNMP query to all the devices.

Does anyone have any thoughts or suggestions on this topic?

Cheers

Troy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.oetiker.ch/pipermail/mrtg/attachments/20150131/2f9a4cae/attachment.html>