MLM is not a substitute for NetView but an adjunct to it. Whether it is
suitable or not for the purpose you want, only you can decide.
MLM only discovers devices in one subnet; it's own. That's why you would
need one for each subnet of the customer's network if you want to use MLM
discovery. it is possible to manually add a few devices in another subnet
to an MLM for monitoring, but there is no way to get it to automatically
monitor or discover more than its own subnet. It is not a tiny NetView.
It's original design purpose was to monitor remote locations connected to
the central NetView by slow WAN links which could not sustain the constant
ping or SNMP traffic that Netview monitoring requires.
MLM's communicate with NetView via SNMP sets and gets, and also traps. You
have to have both the read and write community names correct in
ovsnmp.conf. You can also configure the trapd destination table in the MLM
to threshold on outside traps. Once netmon finds an MLM using an SNMP get
with the read community string, it send an SNMPset to that MLM telling it
to start reporting the discovery and monitoring data it has to the central
location. The nodes the MLM has discovered will get added to NetView
topology provided that netmon can ping them to verify their authenticity.
Combining two distinct customer networks under one NetView will only work
well if they have separate and distinct addressing schemes, since
duplicates are not allowed, unless you also deploy CNAT gateways between
them to do translations.
You might profitably spend some time reading in the MLM doc before making
your deployment decision. You might also deploy a test one and play with
it a little too. That's easy enough to do.
James Shanks
Level 3 Support for Tivoli NetView for UNIX and Windows
Tivoli Software / IBM Software Group
|