nv-l
[Top] [All Lists]

Re: [DEV] Problems with sending traps to NetView R4.1

To: nv-l@lists.tivoli.com
Subject: Re: [DEV] Problems with sending traps to NetView R4.1
From: SHANKS@TIVOLI.COM
Date: Wed, 28 Oct 1998 17:16:53 -0500
Reply-to: Discussion of IBM NetView and POLYCENTER Manager on NetView <NV-L@UCSBVM.UCSB.EDU>
Sender: Discussion of IBM NetView and POLYCENTER Manager on NetView <NV-L@UCSBVM.UCSB.EDU>
Hello Richard -

I was just asked an hour ago to be involved in your problem record.  I will
be sending you  a note off-line to answer your questions.  But first, I
want to say, that if you (or anyone else for that matter) wants to develop
code for sale to work with NetView (or any other Tivoli product) then you
should become a member of the Tivoli Partners Program and get direct  input
into development.  Asking questions via the Support route will work
eventually, but you will not get any guidance on coding or future
development, and worse yet, your code will not be certified as "Tivoli
Ready" when you market it.

I would suggest that the first thing you want to do is to send a note to
Will Mason, the NetView Product Evangelist, and see if he can help you get
into the program and get the contacts you need.   I will copy him on this
response.

James Shanks
Tivoli (NetView for UNIX) L3 Support



Richard Muise <Richard.Muise.rmuise@NT.COM> on 10/28/98 03:15:08 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)
Subject:  [DEV] Problems with sending traps to NetView R4.1





I've contacted the Tivoli support line about this a week ago, and still
don't as yet have a good answer from them, so i thought i'd pose these
two questions to the list. Included is the original message i sent to
Tivoli support line after initial phone conversations.

=======================================
Good Morning.

BackGround:
  We (at Nortel Networks) are bundling a software package called OMS-P
with our ATM Passport switches. We use NetView R4.1 as the base. One
part of the program registers with trapd to recieve all alarms from the
switch, which we reformat into an English readable alarm. This
reformatted alarm is then injected back into trapd where it is sent to
the event browser.

Goals:
1) To be able to set the severity of the trap when we reformat the trap.
The traps (from our software) are all coming out as 'indeterminate'. We
would like to display the trap using a severity level set at the time we
inject the reformatted traps into trapd. This involves finding a way in
software to override the trapd.conf severity level.
2) When we select 'Highlight node on map', this operation fails. We
believe this is because the trap is using the IP address, and we would
like to set the source in our software to the OVWDB selection name.

How we got where we are now:
1) We receive all switch alarms in a program that reformats them. We had
added the oid for the reformatted traps to trapd.conf, which also sets
the default Severity to 'Indeterminate'.
Here is the trapd.conf file parts that are relevant:

#display categories:
CATEGORY 10 "Magellan Passport Events"
#enterprises:
magellanPsm {1.3.6.1.4.1.562.2.3.3.1}
# User added trap formats:
psmPam_Events {1.3.6.1.4.1.562.2.3.3.1} 6 17 V 1 0 "Magellan Passport
Events"
"$3" Passport $8 $5 type:$6 cause:$7 time:$4 cid:$10 comment:$9
SDESC
EDESC

** Note that the severity is set to '1' (indeterminate).

We want to override that severity in the trap at the time we reformat
the trap. So if a critical error comes in from the switch (which are
Passport switches if you were familiar with our ATM product line), we
don't want it to display as indeterminate, but rather as critical in the
event browser.

We tried using information from the NetView R4.1 Administrators
Reference (page 236, trapd.conf). At the top of that page are 4 OIDs
that we could use to set extra information in the reformatted trap. We
thought that .1.3.6.1.4.1.1.1.5.0 might have the severity, but it
doesn't seem to have any effect if we set that OID to the severity level
in the trap.

Please note that we are doing the exact same thing in HP OpenView, and
we can set the severity in the trap to override the trapd.conf
information by setting the OID .1.3.6.1.4.1.11.2.17.2.2.0 as described
in the HP SNMP Developers Guide and Reference, page B-121 (trapd.conf).

So, can we set this information when we run the reformatting program?
The
program runs as a daemon, and this severity should be set by it, and not
by having the user overriding using the NetView user interface.

[addendum. Tivoli support suggested two solutions so far. The first is
to use addtrap to reconfigure trapd.conf as we receive each trap
(totally unacceptable solution) and the other was to add an enterprise
oid for each severity we would have, and have the reformatting program
send the trap as the matching oid (we'll use this solution if there
isn't a way of doing this that is similar to OV).]

2) The second problem that seems related involves using the 'Highlight
node on map' command in the event browser. If we select a trap, we would
like to be able to highlight the node that sent the trap. Right now,
this operation fails for our own 'Magellan Map'. I think it is because
the IP address (or hostname if resolved) is being displayed in the trap,
whereas the actual object name in the map is different (it's not an IP,
but rather a string that is our OVWDB selection name, for example
'EM/NODER90'). Here is an example of the reformatted trap (it uses the
formatting information from trapd.conf included above):

Web Oct 21 13:04:28 1998   47.236.4.133   V  "Lp/1 V35/2" Passport
700112001 CLR type:communications cause:dteDceInterfaceError
time:1998/10/21 12:49:13 comment:blah blah blah

Everything after the IP address is information from our trap reformatter
program. However, since the node name is actually 'EM/NODER90', it can't
find that object on the Magellan Map, because the node is not called by
it's IP address (47.236.4.133). We bind the name 'EM/NODER90' to the OID
.1.3.6.1.4.1.1.1.3.0, as described on page 236 of the Administrators
Reference (same page as noted in problem #1). But the trap still shows
the IP instead of 'EM/NODER90'.

Are we using the OID correctly, or is there a different OID? As with the
first problem, this works fine on OpenView, but with a different HP OID.
The selectionName (using the above OID) is bound to the first varBind of
the reformatted trap.

Thank you for helping with this problem.
-r.

Richard Muise
software developer,
Open Management Systems-Passport,
Nortel Networks
613-763-6981
rmuise@nortel.com

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

Archive operated by Skills 1st Ltd

See also: The NetView Web