nv-l
[Top] [All Lists]

Re: Polling Problems

To: nv-l@lists.tivoli.com
Subject: Re: Polling Problems
From: James Shanks <James_Shanks@TIVOLI.COM>
Date: Tue, 7 Dec 1999 13:06:53 -0500
You are under a serious misconception.

You only get a Node Down (or Interface Down) once and then after that, only if
it goes back up and fails again.  Once the status has been recorded as "down" in
the topology database (and echoed to the object database, you will not get so
notified again.  You are notified of status changes, not polling cycles.

Having said that there was a change in 5.1.2 which resulted in a new Node Down
being issued every time you did a demand poll on a failed device.  But this was
a dramatic departure from previous releasees and several customers complained,
so it will be removed from the code with the next round of maintenance and then
NetView will go back to functioning as it always has, as I just stated in the
first paragraph.

If you really want to watch polling activity then you should run the netmon
trace.  I'll wager you will be surprised at what is going on.

The name of the file where you put a ruleset that you want to run in the
background (that is, by a daemon and not by user in a dynamic workspace) is
/usr/OV/conf/ESE.automation.   But your ruleset should be modified before being
included.  It sounds to me as though you have a Forward Node in it since it is
sending events to the display.  But the daemon (actionsvr) does not have a
display, and if you send him the raw event it will just sit on his queue.  If
you send enough events, his queue will fill up, and he will stop processing, and
then nvcorrd will back up because he cannot send any more to actionsvr.
Eventually you could hang the whole events system back to trapd.    So take the
Forward out of your ruleset.  It should just end with the pager node when you
run it in the background.


James Shanks
Tivoli (NetView for UNIX) L3 Support



"Clay David (rti1dwc)" <rti1dwc@ISMD.UPS.COM> on 12/07/99 12:52:13 PM

Please respond to Discussion of IBM NetView and POLYCENTER Manager on NetView
      <NV-L@UCSBVM.UCSB.EDU>

To:   NV-L@UCSBVM.UCSB.EDU
cc:    (bcc: James Shanks/Tivoli Systems)
Subject:  Polling Problems




I have created a "test" ruleset for paging me when a certain device is no
longer reachable on my map. After the first time netview recognizes the
device is down it send the page to me like it should but after the first
time it never seems to send me a page again nor update my dynamic workspace
event window. It almost seems like it doesn't poll this device after it
recognizes the device is down. Does anyone have an idea what the problem
might be? I assumed netview will poll all devices every 5 minutes so in my
test case I was expecting to receive several pages for this device since I
left it in a down condition. Also does anyone know the name of the file
where you can put a new ruleset you have created so that it automatically is
started when netview is started. The catch here is that in my ruleset I am
forwarding events to the event window for paging purposes when some of our
critical nodes go down. One other question--How does one add an interface to
a router object that wasn't discovered properly? Thanks again.

Dave
xoxoxoxoxoxo


<Prev in Thread] Current Thread [Next in Thread>

Archive operated by Skills 1st Ltd

See also: The NetView Web