Submitted by: Paul Winkeler Paul Winkeler
This ZenPack consists of an rrd template called Barracuda which defined at /Devices/Network. You'll want to bind it to your Barracuda device which presumably sits somewhere below that point.The template contains performance graphs for the inbound queue and the number of:
- bad recipient
All this information is gathered from the Barracuda device by retrieving the contents of the stats.cgi page over an https connection with the "wget" utility. The output of this http GET command is an XML document which is then parsed by some Perl code called "check_cuda" which is provided as a datasource inside this ZenPack / template. The nice thing here is that by not using SNMP, this ZenPack will work for collecting data on even the 300 series (you need to purchase at least a Barracuda 400 for SNMP). It does its monitoring by watching the trailing 24 hour volume of messages and also the inbound message queue size in separate graphs. I only set a threshold on the inbound queue size and I haven't monitored enough of these units yet to know what is reasonable threshold but 45 gets me just a few warnings a day at a site that sees 100k+ messages a day.
If you have any questions, please feel free to contact me at firstname.lastname@example.org
Zenoss Version: 2.2
- The datasource is a Perl script called check_cuda. It uses generic Perl code but does rely on XML::Simple which you may have to install
- The PERL script in turn uses wget to retrieve the Barracuda's basic statistics page; you may need to install it
- The rrd Template assumes your Zenoss installation is at /usr/local/zenoss, you may need to change that assumption
- Finally note that the template associated with this ZenPack defines it self under /Devices/Network so your Barracuda device should be located somewhere at or below this point.
Trac tickets: http://zenpacks.zenoss.org/trac-zenpacks/report/1