nv-l

Re: AW: Netview MLM Discovery

2000-06-07 03:20:10
Subject: Re: AW: Netview MLM Discovery
From: "Gavin Newman" <NEWMANGJ AT banksa.com DOT au>
To: nv-l AT lists.tivoli DOT com
Date: Wed, 07 Jun 2000 16:50:10 +0930
Uwe

Thanks for such a quick response.....

The snmp read/write communities are fine, as evidenced by the fact that the 
trap destination table in the MLM on the Solaris box is correctly updated by 
the Netview process on the AIX box when the MLM is demand_polled from Netview. 
I can also browse the MIB tree on the MLM system OK from Netview.

The problem only appears when the discovery mode is set on in the MLM either 
via smconfig on the MLM host or via the -z parm on Netviews netmon daemon. Once 
that has been done then snmp connectivity is lost to the midmand daemon.

My understanding is that the MLM related snmp traffic is accepted by the snmpdx 
daemon on port 162 and then redirected to midmand on port 1668 by the snmpdx 
daemon. This is where it seems to be breaking down.

I took process lists before and after the change and also made the change early 
one morning so I could see what files had been changed on that day (using the 
find / -mtime 0 -ls command). The only file changes were the 
smMlmCurrent.config file which had the changes related to the new midmand 
configuration. The change in processes was that midmand spawned the child 
process that runs /usr/sbin/arp under a korn shell (sh -c /usr/bin/ksh -c 
"/usr/sbin/arp -a 2>>/dev/null). Apart from that the system is unchanged.

My suspicions are with the new arp process but I can't see how its getting in 
the way.

Cheers - Gavin

>>> Uwe Richter <Uwe.Richter AT synthesis DOT de> 07/06/2000 16:29:00 >>>
Have you READ/WRITE-Community in ovsnmp.conf and snmpd.conf.

Uwe

 ----------
Von:  IBM NetView Discussion
Gesendet:  Mittwoch, 7. Juni 2000 07:56
An:  URICHTER; nv-l
Betreff:  [NV-L] Netview MLM Discovery


I have Netview 5.1 running under AIX 4.3.3 and Netview MLM 5 running on a
Solaris 7 box with the appropriate SEAs installed.
Netview can see the MLM and the C5d daemon has created the MLM Managers   
group
on the Netview GUI.
Traps generated through the MLM (via snmptrap) appear in the Events list   
for
Netview.

So far - so good.

If I try and get my MLM to discover nodes in its network either via the
netmon -z option or by using smconfig on the MLM platform to change the   
MLM
config Netview will mark the MLM as being down. Trying to run smconfig on   
the
MLM box then fails
with an snmp community name error.
Attempts to stop and restart the midmand daemon using smmlm fail with an
snmp_connect() error in the midmand.log file showing that it cannot   
establish
a listening service on the midmand port.
If I stop and restart the snmpdx daemon and then restart the midmand   
daemon
using smmlm -i (to ignore the restart file) I can then use smconfig to   
look
at the MLM config. I then issue a demand poll from the Netview GUI to the   
MLM
and it is again
marked as "up" but, of course, is still not discovering!
Has anyone ever seen this before?

Gavin Newman



**********************************************************************
   *****   IMPORTANT INFORMATION    *****
This document should be read only by those persons to whom
it is addressed and its content is not intended for use by
any other persons. If you have received this message in
error, please notify us immediately. Please also destroy and
delete the message from your computer. Any unauthorised form
of reproduction of this message is strictly prohibited.
Bank SA is not liable for the proper and complete transmission
of the information contained in this communication, nor for any
delay in its receipt.
**********************************************************************
_________________________________________________________________________
NV-L List information and Archives: http://www.tkg.com/nv-l 

_________________________________________________________________________


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