For those of you who file away the odd-ball stuff for a rainy day, here is
the
solution to last week's puzzler...thanks to Lynn in Support!
The answer is: use loadhosts! Something I never use, being a big proponent
of
natural discovery as a means of identifying errors in (!) subnet
masking....
Loadhosts lets you specify the mask, so the device gets added to the right
subnet. Netmon has not the wherewithal to move it back, or even complain.
So we used a Collection to identify the nodes to be moved; we used
nvdbformat to
generate a report listing the hostnames and the TopM Interface list (to get
the
loopback address) of the members. We sorted that by IP address to make
chunks by
subnet mask. We used the chunks to run ovtopofix -r to delete the nodes,
and
loadhosts to add them back again. A pretty good excercise for the student.
Cordially,
Leslie A. Clark
IBM Global Services - Systems Mgmt & Networking
===================================================================
This is a new one on me. Is anyone out there running with unadministered
masks on their hubs and switches?
I have a customer who only administers them on the routers because they use
proxy arp and therefore don't
need masks on end nodes. Of course Netview is drawing every hub, switch,
and server in the whole network
on the same class-B subnet. This makes it hard to associate a device with a
section of the network, to say the least.
So has anyone come up with any clever solutions, or did you just break down
and apply the masks?
Cordially,
Leslie A. Clark
IBM Global Services - Systems Mgmt & Networking
|