I'm not sure if this got answered today, and if I wait another hour or so I
could check the
archive (isn't that great!) but I might not know the answer in an hour.
First, you want to make sure that the event, as it shows in the display, is
shown as coming
from a name. It comes in from an address, but gets resolved if possible, to
a name.
If that node is on your map with a name, and the trap shows as from an
address, they will
never match. So you may need to add names to your chosen method for name
resolution.
If you can select the event and do 'Options..Highlight Node on Map', then
you are doing it right.
Often nodes are discovered by the lan address because that's what the name
in the seedfile
resolves to, but then the traps come in from a serial link which is not
named. Name it.
Now, in the ruleset: For agent traps, you want to match Origin in the trap
with membership
in the collection. For Netview events, it can be either variable 2, or
Origin, which is
artificially set to be the same as variable 2 (since the real origin is the
netview server).
I hope that is clear.
Cordially,
Leslie A. Clark
IBM Global Services - Systems Mgmt & Networking
Detroit
=============================================================
Hello ,
i am having trouble defining a ruleset in which i would like to check
whether an alert is coming from a specific list of hosts ( which are in a
collection ).
I do a collection query in the ruleset , but what should I put as the
'object ID source ' so that netview would compare the ip-adress of the
sending host with the adresses in the collection ?
|