Re: FW: Problem for represent meaningful information in action

1999-08-12 08:13:43
Subject: Re: FW: Problem for represent meaningful information in action
From: James Shanks <James_Shanks AT TIVOLI DOT COM>
To: nv-l AT lists.tivoli DOT com
Date: Thu, 12 Aug 1999 08:13:43 -0400
Maggie -

First, NetView 5.0 is not supported.  You MUST install NetView 5.1 and then
apply maintenance to 5.1.1 -- in fact 5.1.2 will be out shortly.  I would also
call Support and get any emergency fix packages that you need for 5.1.1. to tide
you over until 5.1.2 comes out.  There is no point in installing MLM at the
level of code you have, unless you then plan to do it all again later.

(A) There are only going to be as many $NVATTR_x variables as there are
variables in the trap.  If the trap only has 15 variables, then only $NVATTR_1
through $NVATTR_15 will be provided.  The others will be NULL.   I think you are
still ignoring my explanation and trying to find things in the trap which are
not there.  You have to find out what all the variables mean and supply your own
formatting in the action node.

(1)  MLM and NetView for UNIX should not be installed on the same box.  So there
is no reason to change the trap port, but if for some strange reason you insist
on MLM and NetView for UNIX on the same box, then midmand must get a new port.

With NT, MLM for NT and NetView for NT can be installed on the same NT box.
This configuration is called an "attended" MLM and you will find information
about it in the \usr\ov\doc\netview_mlm_readme.htm  file.  But this is only
available in NetView V5.1 and later.

(2) When netmon is started it automatically goes looking for MLMs.  When it
finds one, it updates the object and topology databases and then queries the MLM
to off-load polling if you have configured it for that.  From then on it just
status polls the MLM.  If the MLM fails to respond, then netmon automatically
takes over polling of the devices MLM was polling.  Please read all the
documentation on this that you can before implementing this function.

James Shanks
Tivoli (NetView for UNIX) L3 Support

"Sin Wai Yee, Maggie" <mwysin AT KCRC DOT COM> on 08/11/99 11:34:20 PM

Please respond to Discussion of IBM NetView and POLYCENTER Manager on NetView

cc:    (bcc: James Shanks/Tivoli Systems)
Subject:  FW: Problem for represent meaningful information in action box of
      ruleset & Concerns about using MLM

> I have tried to display up to $NVATTR_15, but still meaningless.
> Therefore, I stop at there.
> Thanks for your advise, I will try all the $NVATTR and take a look on the
> programmer guide first.
> Can I ask other question ?
> I am using 5.0 Netivew for AIX for whole network. We have bought several
> TME 10 Netview for NT for distributed monitoring.
> The idea is
>       several big site ( out of headquarter ) will install MLM ( NT ) for
> monitoring and then send an important information to headquarter ( where
> we have Netview for AIX ).
> May be I am not understand MLM in deep, I have several concerns below :
> Do anyone using MLM, please kindly make you advice :
> Question 1 :
>       If the headquarter one implement MLM also ( MLM take a
> responsibility for monitoring and discovering ) , default snmp trap port
> should be changed. If any MLM in local node or in remote node have failed,
> whether it can be a backup or not ( due to the change of the snmp trap
> port ).
> Question 2 :
>       If the headquarter one not implement MLM, is it can properly receive
> an important message from MLM of remote node. And how can I stop the
> netview ( central one ) to poll the remote area ( which suppose to be the
> responsibility of MLM of remote node.)  And for this case, the central
> netview can act as a backup of MLM of remote node.
> Thanks
> -----Original Message-----
> From: James Shanks [SMTP:James_Shanks AT TIVOLI DOT COM]
> Sent: Wednesday, August 11, 1999 8:50 PM
> To:   NV-L AT UCSBVM.ucsb DOT edu
> Subject:      Re: Problem for represent meaningful information in action
> box of ruleset
> Maggie -
> Perhaps it is just me but I am having a very hard time following what you
> are
> trying to tell me.  Perhaps the problem is that you are expecting to get
> something out of the ruleset action that is not there.  But without seeing
> your
> action command for sure, I cannot tell.
> You keep jumping back and forth between how the trap can be displayed in
> trapd.log or in the event window and what you get out of the action node
> of a
> ruleset.  These are two entirely different things.  You are correct that
> loading
> a MIB and adding traps to trapd.conf has no effect whatsoever on what you
> get
> out of the ruleset.  Did you expect that it would?
> In a ruleset action, you must do all the formatting.  In trapd, for the
> most
> part, you need only tell trapd what you want, by configuring trapd.conf.
> I hope
> the difference is clear.  All that gets passed to the ruleset is the raw
> trap.
> It is not formatted.
> Now we are back to the beginning.  You MUST  make the effort to figure out
> what
> the trap variables in the 8260 mean.  I cannot do it for you.  Once you
> know,
> then you can fix your action node script to send a meaningful mail
> message.
> Perhaps you can use the trap formats in trapd.log and the event window to
> help.
> If you make the event log message read "$*" it will show you all the
> variables
> with what whatever MIB definitions you have loaded.  Then you can use that
> information to add your own formatting to mail message.  But if all you do
> is
> echo the variables, $NVATTR_1 to $NVATTR_50, then all you will get is
> whatever
> was in that position in the trap, and there will be no formatting done.
> What
> you are seeing is not only normal but expected.
> There are NO APIs to read trapd.conf and extract the message formatting
> but you
> can do that in a script if you like. I am afraid you are on your own here.
> All NetView APIs you may use for whatever purpose are documented in the
> NetView
> Programmer's Reference.   Hints on how to  design programs are discussed
> in the
> NetView Programmer's Guide.
> Hope this helps.
> James Shanks
> Tivoli (NetView for UNIX) L3 Support

<Prev in Thread] Current Thread [Next in Thread>
  • Re: FW: Problem for represent meaningful information in action, James Shanks <=