This is a problem with thresholds in general. Most thing that do threshold
monitoring don't
reset status at startup. The threshold has to be passed again, and then
re-armed, before
the status gets reset to ok. This is also a problem for nodes that have
passed a threshold
while you had the system down. You might not find out about it until the re-arm
point has
been passed, and then the threshold gets passed again. Some applications
address this by
just not re-arming at all. For instance Nways PSMs will set the status of the
node when a trap
comes in, and you are supposed to go investigate (eg Monitor..Events...Current
Events for the
selected node) and then ack/unack or manage/unmanage the node to get the status
reset to IP-only.
The Agent Policy Manager took a different approach, creating its own objects to
reflect
conditions. Another factor might be the setting of the option for the map
(File..Describe.. ) that
controls whether IP status overrides these things. You can also take a look at
how many
events you are loading, in /usr/OV/app-defaults/Nvevents and Nvela. Maybe you
can afford
to keep enough (memory-wise) that the related event will still be loaded when
you restart
the map. Even if you can't load enough, you could still do a grep on the
/usr/OV/log/trapd.log
file to see what had come in for that node.
So, you need to figure out which products or components are doing the
thresholding and
dig into how they are supposed to work.
Cordially,
Leslie Clark
IBM Global Services - Network & Systems Management - Detroit
----------------------------------------------------------------------------------------------------------------------------------------------
I have a netview v 4.1.4 for AIX and have the following problem with
the control desk:
When I receive a trap (threshold trap for instance), the icon turns to red
and a message is displayed on the control desk.
Then I logout and stop all netview daemons (for week-end maintenance).
When I login back again, the icon is still red but the message disappeared.
This is quite confusing because I can't find out which daemon causes this
problem.
Does anybody have and idea about this ?
Cheers,
Eric.
-------------------------------------------------------------
Eric Filippi (efilippi@eurhone.e-mail.com)
Eurhone
IBM Global Service Ouest
Lyon
FRANCE
-------------------------------------------------------------
|