[Top] [All Lists]

RE: [nv-l] internal trap formatting

To: nv-l@lists.us.ibm.com
Subject: RE: [nv-l] internal trap formatting
From: "Evans, Bill" <Bill.Evans@hq.doe.gov>
Date: Wed, 12 Apr 2006 09:56:51 -0400
Delivery-date: Wed, 12 Apr 2006 14:57:46 +0100
Envelope-to: nv-l-archive@lists.skills-1st.co.uk
Reply-to: nv-l@lists.us.ibm.com
Sender: owner-nv-l@lists.us.ibm.com
Thread-index: AcZeM0C/dI61KVBHTtSrrEC/Y8igCQAA3a1w
Thread-topic: [nv-l] internal trap formatting
That same little table of SystemName.InstanceNumber=InstanceDescription
could be copied to the TEC system and the rules there do the resolution.
This assumes the table doesn't change very often.  It's about the same
amount of initial work to implement but more efficient in operation.
Mention of the tecad_nv6k.cds file would imply Praveen uses TEC.   

As I remember it, we in NetViewLand rejoiced when IBM acquired Tivoli so
that the TEC Rules could fulfill our need for event correlation. This is
one case where those rules can do the job more efficiently.  

Of course, if TEC is not involved or the concern is for the NetView
Console displays, the solution James mentions is the best available. 

Bill Evans

-----Original Message-----
From: owner-nv-l@lists.us.ibm.com [mailto:owner-nv-l@lists.us.ibm.com]
On Behalf Of James Shanks
Sent: Wednesday, April 12, 2006 9:16 AM
To: nv-l@lists.us.ibm.com
Subject: [nv-l] internal trap formatting

I do not see any simple nor easy way to achieve this.  The trap is sent
by snmpcollect, who collects by interface number, not by interface
description.  As currently designed, he does not gather that
information, so it is not included in the trap.  Thus there is no mere
manipulation of the trap format that you could do to add it.  Rather,
you would have to design a mechanism of your own to obtain it.

You could, for example, write a script to be launched from trapd.conf as
a hidden application every time this trap is received, and have it
extract the hostname and interface number from the incoming trap, and
then do an snmpget for .<instance> , which would
return the ifDescr for that interface.  Then you could use that with the
command to re-issue a trap of your own containing this new value.   But
you can see this would be a great deal of effort.   As an alternative,
perhaps, you could look up the interface names in advance, and build a
table in a file, and have your script just get the matching name from
there.  But you'd still have to re-issue the a trap of your own with
this new information added.

Anyone else have any ideas?

James Shanks
Level 3 Support  for Tivoli NetView for UNIX and Windows
Tivoli Software / IBM Software Group

"Praveen Kumar" <Praveenkumar.nair@ustri.com>
Sent by: owner-nv-l@lists.us.ibm.com
04/12/2006 07:36 AM
Please respond to

Hi all,

I am running netview 7.1.4 on win2003.

Some of the SNMP collect status events generated by Netview has some
predefined format message. By looking at the tecad_nv6k.cds file, I
could see this description is mapped with a trap variable "nvEventDescr"
(Respective OID is Since this preformatted
event description holds some value like instance number rather than the
description of that instance, this is not easily understandable.

For instance,

I have two processors in a machine and started the collection for Max
CPU utilization. I set a threshold of 75%. When Event is triggered the
resulted event message comes as "<ExpressionName> <Instance Number>
threshold trigered (>=75.00): 80.00". Here what I prefer is Instance
description rather than Instance number, which would be meaningful as
well as easily can be made out.

Would someone help me in resolving this issue?

Thanks in Advance,

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

Archive operated by Skills 1st Ltd

See also: The NetView Web