I'll take a shot at this. If anyone
else has an opinion, please chime in.
>1.
Once a node is discovered and managed, if the underlying data flow,
paths, hardware, etc change, does the end Node need to be rediscovered
(deleted and discovered) in the NV DB?
No, I don't think so, as long the IP
address of the main interface remains the same, and as long as NetView
can get to the device, and it still responds to SNMP. Configuration
polls should take care of everything else that I can think of.
>2.
Is it possible in NV to not discover through the cloud to get to
the end Node, but instead create symbols/objects for the end nodes and
demand poll them and they be added to the NetView DB, thus circumventing
the cloud all together?
This is done by turning off discovery
and using loadhosts. But you need the correct subnet mask, and the
device has to ultimately be accessible through the cloud by ping/SNMP,
else it will be blue forever.
James Shanks
Level 3 Support for Tivoli NetView for UNIX and Windows
Tivoli Software / IBM Software Group
"Allison, Jason (JALLISON)"
<JALLISON AT arinc DOT com>
Sent by: owner-nv-l AT lists.us.ibm DOT com
03/23/2004 04:00 PM
|
To
| "nv-l (nv-l AT lists.us.ibm DOT com)"
<nv-l AT lists.us.ibm DOT com>
|
cc
|
|
Subject
| [nv-l] Discovery and network
migration, couple questions |
|
Going to try to not confuse with too many
details.
We are migrating underlying infrastructure
and I am working on the procedures to migrate currently discovered and
managed nodes as well as the procedures for discovering and managing new
nodes. Good news is neither the NMS or the old end Nodes changing
addresses. For lamens sake we have:
NMS -> <cloud A> -> Managed Node
Migrating to ...
NMS -> <cloud B> -> Managed Node
Our paradigm is to manage through cloud A,
manage end Node, unmanage and hide cloud A on map.
1. Once a node is discovered and managed,
if the underlying data flow, paths, hardware, etc change, does the end
Node need to be rediscovered (deleted and discovered) in the NV DB?
Our new Cloud B border router has an access
list to not allow SNMP traffic (any traffic) to the NMS originating from
within the cloud, HOWEVER it will allow traffic originating from the end
Node. This breaks our current set of procedures to manage through
the cloud to get to the end Node.
2. Is it possible in NV to not discover
through the cloud to get to the end Node, but instead create symbols/objects
for the end nodes and demand poll them and they be added to the Netview
DB, thus circumventing the cloud all together?
Thanks in advance,
Jason Allison
Principal Engineer
ARINC Incorporated
Office: (
|