[TriLUG] OpenNMS and mapping

[gregbrown at mindspring.com]

Yeah, we've got sniffers by the boat-load.  Literally.  At this time there are two networks operating side-by-side, the old DEC and Cabletron network with the 16 bit mask and the new Cisco-based "routed" network with the 24.  When we migrate a building we physically move connections from the old network to the new, sexy, 10-gig network.   So, eventually, all that will be left over on the old network are the gigaswitches, cabletron FDDI attached hubs and other edge devices.  So if I look at a map and nothing is "hanging off" any edge devices I'm in good shape and I can turn off the old backbone FDDI connection points.  Problem is when I do nmap scans of the "old" network I get way more IP addresses out there then the amount of core and edge network devices.  So there are things out there..  Somewhere.  I'm trying to group them by vendor ID of the MAC addresses to determine what is likely to be a user workstation or a network attached printer but that is tedious.  And it still doesn't tell  me where the devices are located (on which campus).  

Sniffers might work too, though.

Greg

-----Original Message-----
From: Shane O'Donnell <shaneo at nc.rr.com>
Sent: Nov 17, 2004 2:06 PM
To: 'Triangle Linux Users Group discussion list' <trilug at trilug.org>
Subject: RE: [TriLUG] OpenNMS and mapping

So is anything changing other than the netmask?  If not, I don't think your
map is going to help.  You need a sniffer.  That or I'm missing something.

Surely this "Good Sized Kompany" has sniffers in remote environments -- at
least it's major ones, right?

Shane O. 

-----Original Message-----
From: trilug-bounces at trilug.org [mailto:trilug-bounces at trilug.org] On Behalf
Of gregbrown at mindspring.com
Sent: Wednesday, November 17, 2004 1:22 PM
To: Triangle Linux Users Group discussion list
Subject: RE: [TriLUG] OpenNMS and mapping

Agreed, maps are evil - except when they are useful in a point situation
like this.  Here's what is going on:

1 Company X, who I can't actually name for some legal reason, is migrating
away from a large, flat class B network to a new class B address range that
is routed with a 24 bit mask.
2. There are multiple campuses seperated by large geographic areas but these
multiple campuses are still falt - i.e. the "old" class B has a 16 bit
subnet mask and are all connected via long-range FDDI.

This all means that it is nearly impossible to tell if a given IP address is
in, say, zebulon, RTP, Durham, or any one of a zillion, it seems, other
locations.

The reason I want THIS map is for only one reason:

to confirm that any one campus has fully migrated from the old network to
the new network.   The whole gigantic FDDI ring is going to start breaking
up campus by campus and I want to visually check to make sure there are not
stragglers prior to turning the power off to the FDDI switches.

Going through DEC's bizarre interface to try to see where a MAC address is
located is just getting to costly in terms of time so I've opted to go the
map route in this case.

Unless anyone has any other suggestions.....

Greg



-----Original Message-----
From: Shane O'Donnell <shaneo at nc.rr.com>
Sent: Nov 17, 2004 11:43 AM
To: 'Triangle Linux Users Group discussion list' <trilug at trilug.org>
Subject: RE: [TriLUG] OpenNMS and mapping

Both NetSaint/Nagios and OpenNMS' "aftermarket" mapping bolt-on are, to
reiterate the sentiment of the all-knowing Jon Carnes, kludgey at best.

Depending on what you are trying to do and what level you are trying to map
-- yet another reason why maps are a bad idea in the first place, but I
digress -- checkout NetDisco (http://www.netdisco.org/)  They have a running
demo on their site and a screenshot available of UCSC's network, which
scales beyond the usefulness of maps -- yet another reason why maps are a
bad idea in the first place, but I digress...

Network management is about information collection (active and passive) and
display for someone who requires that information to do their job.  Maps are
for managers.  Unfortunately, managers also have purchasing budgets, so
tools that provide maps get bought and sit on the shelf while the people
that need the tools use OpenNMS, Nagios, MRTG, and the like to get their
work done.

Don't get me started...

Shane O. 

-----Original Message-----
From: trilug-bounces at trilug.org [mailto:trilug-bounces at trilug.org] On Behalf
Of Jon Carnes
Sent: Wednesday, November 17, 2004 11:13 AM
To: Triangle Linux Users Group discussion list
Subject: Re: [TriLUG] OpenNMS and mapping

On Wed, 2004-11-17 at 11:02, gregbrown at mindspring.com wrote:
> Does OpenNMS graphically map out networks it manages?  If OpenNMS does not
does anyone know of a OSS tool that would create graphical maps of a network
via autodiscovery?
> 
> Greg
> 
I run an older version of OpenNMS and it doesn't automap. Jaimie Livingston
demoed the latest version of NetSaint (not called that anymore) and it *did*
automap the network. Automap was cool, but a bit kludgey looking.

Jon Carnes

-- 
TriLUG mailing list        : http://www.trilug.org/mailman/listinfo/trilug
TriLUG Organizational FAQ  : http://trilug.org/faq/
TriLUG Member Services FAQ : http://members.trilug.org/services_faq/
TriLUG PGP Keyring         : http://trilug.org/~chrish/trilug.asc

-- 
TriLUG mailing list        : http://www.trilug.org/mailman/listinfo/trilug
TriLUG Organizational FAQ  : http://trilug.org/faq/
TriLUG Member Services FAQ : http://members.trilug.org/services_faq/
TriLUG PGP Keyring         : http://trilug.org/~chrish/trilug.asc



-- 
TriLUG mailing list        : http://www.trilug.org/mailman/listinfo/trilug
TriLUG Organizational FAQ  : http://trilug.org/faq/
TriLUG Member Services FAQ : http://members.trilug.org/services_faq/
TriLUG PGP Keyring         : http://trilug.org/~chrish/trilug.asc

-- 
TriLUG mailing list        : http://www.trilug.org/mailman/listinfo/trilug
TriLUG Organizational FAQ  : http://trilug.org/faq/
TriLUG Member Services FAQ : http://members.trilug.org/services_faq/
TriLUG PGP Keyring         : http://trilug.org/~chrish/trilug.asc