Dave is right. I've had good results with predicting failures by catching
flapping. Configure the devices to send link up/down traps to Netview,
usually a core switch is a good candidate. On some devices this is done
at the interface level and on others it is all or nothing. I would not do
this for edge devices, though.
Get fancier.
Make a ruleset to threshold on a link down trap from the same device
for the same interface, say 3 in 10 minutes or something. Add an Action
that calls a little script to generate a custom trap (eg specific trap
1001 in the Netview enterprise) to say that interface x on node y is
flapping. Run the rule in the background (/usr/OV/conf/ESE.automation).
Now you have a meaningful event to take action on, or list in reports.
If you also monitor (with snmpCollect) the error rate (>1%) on those
interfaces, you will get threshold and rearm traps. Taken together, these
events will indicate problems that simple 5-minute pinging will never
show you.
Implementing these functions will be a great exercise in customizing
several areas of Netview.
Cordially,
Leslie A. Clark
IBM Global Services - Systems Mgmt & Networking
Detroit
reamd@Nationwide.
com To: "Donald Mahler"
<dmahler@telcordia.com>
cc: nv-l@lists.tivoli.com
07/11/2002 12:53 Subject: Re: [nv-l] Tivoli
NetView - detecting
PM flapping/flakey links
Configure network device with snmp and forward traps to netview..
"Donald Mahler"
<dmahler@telcordia.com> To:
nv-l@lists.tivoli.com
cc:
bcc:
Subject:
[nv-l] Tivoli NetView -
detecting
flapping/flakey links
07/11/2002 11:18 AM
1) We had a problem the other day in which we did not detect a degrading
link until the customer did (not good). The problem was that link was
degrading... starting to fail... .. Our snmp intf polls of the router did
not see the link drop, and netview pings to the remote side did not fail
either. Now these netmon do a fair amount of retries, and if at least one
works, then netview does not alert. But.... if a ping only works once
out of 5 retries, for a few hours, reason tells you that something is
wrong... sigh... this knowledge is buried within netmon I suppose... and
there is no external "polling log".... How do others detect this sort of
brownout condition?
2) Also, we do not tend to dispatch people in the middle of the night if a
link does down for under 1 minute. But if a link is down/up, down/up,
down/up, down/up, all the night long, we would like to be able to
recognize this condition . Any way to have netview detect this flapping?
or do I have to code this into TEC?
3) or....perhaps I just need to make sure that we are trending and alerting
on iferrors and the like
anyways..... any and all comments welcome
Don Mahler
SAIC
---------------------------------------------------------------------
To unsubscribe, e-mail: nv-l-unsubscribe@lists.tivoli.com
For additional commands, e-mail: nv-l-help@lists.tivoli.com
*NOTE*
This is not an Offical Tivoli Support forum. If you need immediate
assistance from Tivoli please call the IBM Tivoli Software Group
help line at 1-800-TIVOLI8(848-6548)
---------------------------------------------------------------------
To unsubscribe, e-mail: nv-l-unsubscribe@lists.tivoli.com
For additional commands, e-mail: nv-l-help@lists.tivoli.com
*NOTE*
This is not an Offical Tivoli Support forum. If you need immediate
assistance from Tivoli please call the IBM Tivoli Software Group
help line at 1-800-TIVOLI8(848-6548)
|