[c-nsp] using snmp views effectively

[Ray Burkholder]

>Utilisation(interface)
>Errors/Discards(interface)
>CBWFQ policies on those interfaces?


All these statistics are recorded by Cricket.  For alerting, there are a
couple of options that come to mind.  I'm sure others can think of others.

1) Cricket has an event interface.  I havn't used it personally, but
supposedly it might to do the job for alerting.
2) Since the API against RRD files is published, one could run a periodic
scanning script against them to automatically detect values out of range and
report them.
3) Combine a perl script that can unlock the secrets of the cricket config
file with one that can unlock the secrets of the rrd files, with the result
being something similarily effective.

Alternatively, http://torrus.org/ might be another tool that brings all this
together with their threshold monitor.  The website indicates a Debian
package available, so it might now be easy to get up and running.  It looks
difficult to configure.  I havn't figured out if they have a nifty tool such
as the acktomic toolset to automatically create a config file for all
interface types found in a device.

Something nebulous that has been floating in mind is something like a
pattern matching tool that can allow daily cycles but knows when something
deviates from the norm.  I've heard mention of such a thing, but cannot
recollect where to start a search.


-- 
Scanned for viruses and dangerous content at 
http://www.oneunified.net and is believed to be clean.