[c-nsp] OT: Open Source Network Monitor

Mike Butash der.mikus at gmail.com
Wed Sep 6 01:57:05 EDT 2006


Try NMIS for cisco-centric networks, it really is a rather comprehensive 
solution for monitoring everything about a switch or a router.  Anyone 
that runs Cisco should have an nmis install, it blows away just about 
anything commercial for functionality and valuable information.

http://www.sins.com.au/nmis/

It pulls and graphs pretty much everything like memory, cpu, buffers, ip 
stats, and just about anything related to the interface.  Recently they 
added cbwfq stats that do a rather nice job breaking down transmit/drop 
stats per class, and supposedly protocol stats from nbar as well.  I'd 
recommend CVS builds for the cool toys and not be afraid to work through 
30 or so perl dependencies, as their install docs are generally quite 
out of date for what it is going to want to work.  Has some other nice 
toys like reverse-dns generators for interfaces, nice inventory output 
for devices/serial numbers, modem async usage, pix failover/connection 
stats, etc.  It does alerting and thresholding stuff too, but the stats 
on the devices and interfaces eclipses just about anything else for 
being worth the hassle to install.

If you have a ton of chassis switches or tons of 48 port 
access-switches, you need to be mindful of scalability concerns with 
threading and unfinishing cron processes - this will obliterate the 
beefiest box in time with semaphore depletion (at least in linux).  It 
has master/slave functionality to distribute load between boxen, and be 
ready to implement it to keep things happy once you get too big.

Zabbix holds promise as a suitable replacement, as it uses app binaries, 
php, mysql, gd for graphing, and is almost fully capable of having it's 
application tiers load-balanced.  Not nearly as automagical in nature as 
NMIS, but definitely holds promise to scale far beyond NMIS, and can use 
  scripts to perform the modeling ala NMIS for determining what all to 
poll of a known class of devices.

-mb


Max Reid wrote:
>> Nagios doesn't do traffic graphing via SNMP, for example.
> 
> Exactly!
> 
> Common setup is:
> 
> Nagios for Alarms
> 
> Cacti for Trending
> 
> 
> Regards,
> Max
> 
> 
> 
> 
> 
> 
>> On 30-Aug-06, at 6:46 PM, Paul Stewart wrote:
>>
>>> Just curious..:)  Why would you need Cacti if you already are running
>>> Nagios?  Doesn't Nagios graph everything you could need?
>>>
>>> Take care,
>>>
>>> Paul Stewart
>>> Network Administrator
>>> Nexicom Inc.
>>> http://www.nexicom.net/
>>>
>>> -----Original Message-----
>>> From: cisco-nsp-bounces at puck.nether.net
>>> [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Jason Lixfeld
>>> Sent: Wednesday, August 30, 2006 5:04 PM
>>> To: Joseph Jackson
>>> Cc: cisco-nsp at puck.nether.net
>>> Subject: Re: [c-nsp] OT: Open Source Network Monitor
>>>
>>> I would suggest Nagios + Cacti (running that combination currently).
>>> Followed closely by JFFNMS.  Nagios is a very mature product with a
>>> large user/support base and is very flexible with regards to custom
>>> plugins for things like SNMP polling (which is doesn't do natively
>>> (yet??)).  Cacti has matured greatly in the last couple of years
>>> and the
>>> configuration complexities have been greatly reduced from where they
>>> were when the project was in it's infancy.  The two aren't integrated,
>>> but there are some third party plugins and patches for both to try to
>>> integrate the two (needed because Nagios doesn't have graphing
>>> functions
>>> for things like bandwidth monitoring/errors/pps/ etc and likewise,
>>> cacti
>>> doesn't have any sort of robust device monitoring/alerting
>>> capabilities).
>>>
>>> JFFNMS has come along way.  It provides functionality similar to a
>>> hybrid of Nagios and Cacti as it performs both host monitoring/
>>> alerting
>>> and graphing.  The monitoring/alerting functionality isn't as
>>> robust as
>>> Nagios, but the graphing side of it is quite good.  I don't
>>> particularly
>>> care for the UI, but it too has matured over the last couple of years.
>>>
>>> HTH...
>>>
>>> On 30-Aug-06, at 4:05 PM, Joseph Jackson wrote:
>>>
>>>> Hey all,
>>>>
>>>> 	I know this is way off topic but I would like to know any
>>> software
>>>> products you guys can recommend.  We are looking for an open source
>>>> network monitor.  I have tried OpenNMS but didn't like certain parts
>>>> of it so I'd like to hear what you all gotta say.
>>>>
>>>>
>>>> Joseph
>>>>
>>>> _______________________________________________
>>>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>>>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>>>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>>>>
>>> _______________________________________________
>>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>>>
>> _______________________________________________
>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>>
> 
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
> 


More information about the cisco-nsp mailing list