I would guess you have a real network issue every 21 minutes. As Leslie
would say, if NetView is doing something goofy your network is probably
doing something goofy. Some questions I might ask:
Can you see a flood of traps just prior to the time the node downs start?
Maybe a single trap
Do you have RFI turned on and based on your servers location in the network
and where the managed nodes are, could you be seeing your local router go
unreachable? (If RFI is turned off, you would get a flood of node down
traps) (OOPS, version 5.1 netview no RFI - my bad)
Do you have a CPU intensive operation (like SNMP collect) that occurs on a
20 minute interval? Could you be pegging your network interface or your CPU
to the point where pings begin to fail?
Is there another device in the network monopolizing the bandwidth between
you and the nodes that are failing? (Is there a serial link between your box
and the end locations?)
Could you have typo'd the status poll time and made it tenths of seconds
rather than full seconds (like waiting .1 seconds instead of 1 second for
ping responses)
Does pinging the device immediately after you see it turn red, turn it back
to green?
Are you using a ruleset that catches node ups and downs and maybe a pass or
reset on match is hanging up events and letting them roll out all at once?
(Is there anything dependant on 20 minute timers in your rulesets?)
Sorry if these don't sound like a lot of help - but this sure sounds like
there is a real problem in your network.
-----Original Message-----
From: owner-nv-l@tkg.com [mailto:owner-nv-l@tkg.com]On Behalf Of Stover,
Joseph E
Sent: Friday, August 03, 2001 1:09 PM
To: 'IBM NetView Discussion'
Subject: [NV-L] Netview events up/down
I am running netview 5.1 on aix 4.2.1. I realize that this should be
upgraded. The system has been working fine, but recently approximately
every 21 minutes all of the nodes enter the down state, and then 2 mins
later it comes back up. It's rather annoying, I am now trying to figure
out if it's a netview anomoly or if something is wrong with the tcp/ip
stack. During this time, telnet sessions are extremely sluggish and my snmp
scripts (gets/walks/sets..) timeout also.
Has anyone else had this occur? What causes it and or fixes for it?
Thanks
Joseph Stover
Bureau of Technology Services
Pennsylvania State Police
1800 Elmerton Avenue
Harrisburg, PA 17110
_________________________________________________________________________
|