Re: FW: FW: Solstice Enterprise Agents for NetView 5.1.1

1999-08-18 10:41:57
Subject: Re: FW: FW: Solstice Enterprise Agents for NetView 5.1.1
From: Jason Kravitz <Jason_Kravitz AT TIVOLI DOT COM>
To: nv-l AT lists.tivoli DOT com
Date: Wed, 18 Aug 1999 10:41:57 -0400
"Giovannini, Tom" <tom.giovannini AT EDS DOT COM> on 08/18/99 10:19:36 AM

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

cc:    (bcc: Jason Kravitz/Tivoli Systems)
Subject:  FW: FW: Solstice Enterprise Agents for NetView 5.1.1

Tom - there is a Solaris kernel patch which should fix the DMI problem... so
instead of completely disabling DMI... apply kernel patch 106149-07

I am told that we are shipping this kernel patch with future versions of the
NetView kit - otherwise you can download it from Sun...

>What is the best way to disable the DMI process?

see above

>What do I lose by disabling the DMI process?

You don't lose anything by disabling DMI unless your shop relies on DMI for
information - DMI is simply another means of obtaining system information from a
remote machine (disk space, OS level, hardware etc) - I don't deal with DMI too
much and am not sure of the exact specifics.

>Can you elaborate a little on how "1.0.2 seems to operate with NetView
>better than the SNMP agent in 1.0.3"?

The mibii subagent seems to randomly go down but the master SNMP agent remains
up - NetView relies mainly on the mibii subagent for queries so if this is not
running... netmon will not be able to perform SNMP status polls and will go down
- the solution, recycling snmpdx  (/etc/init.d/init.snmpdx stop / start) will
recycle the master and sub agent so this fixes the problem.  The subagent in SEA
1.0.2 does not seem to randomly go down and is more stable.

Strangely, we were seeing a very similar problem with SEA 1.0.1 and it worked in
1.0.2 - I'm not sure what changed in 1.0.3 to reintroduce this problem and we
are not able to recreate consistently enough to raise the issue with Sun


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