nv-l

Re: what's going on...

1998-06-21 15:04:40
Subject: Re: what's going on...
From: Jim Kellock <jkelloc AT IBM DOT NET>
To: nv-l AT lists.tivoli DOT com
Date: Sun, 21 Jun 1998 15:04:40 -0400
Regarding levels for Nways first- Do not try to use ANY mgtapptran ptf
with ANY currently supported level of Nways- either 1.1 or 1.2.x.
As of 1.1, mgtapptran shipped with the Nways product and matched the
PSMs in that product level.  [Nways 1.1 goes out of defect support at
end of June (this week)].

Nways levels- Nways Manager 1.1 contains individual appls. at 3.2.x
              Nways Manager 1.2     "        "        "     " 3.3
              Nways Manager 1.2.1   "        "        "     " 3.3.1
              Mgtapptran is a little different- at 1.2.1 NM its 3.0.1.1

If Red is chewing up CPU, after its initial polling, first check for
orphans children with 'ps -ef | grep R' (note upper case- all of the
kids contain one too).  If there's anybody running with a parent pid of
1 kill it, then do what Leslie says- Tools->Applic.Trans.->Stop, then
same string and Start.  If you get an error on the console when the
daemons start back up, stop them again, do a 'ps -ef | grep lockmgr',
get the pid of that and kill it (he's supposed to be running under init,
so that's OK, but if an mgtapptran session dies unnaturally the lockmgr
seems to have a pid file for all Red instances (one for each NV session)
that doesn't get cleaned up and he gets confused.)  After you kill
lockmgr, do the start again.

If there's still a problem, there may be some database issues, so go
into smit->commo->Applic.Transporter->Maintain and clear the database.
(This has no affect on any Netview database object).  If this fails to
complete, it's because you don't have enought room in /tmp to dump the
database contents.  After it finishes grinding through another tedius
synchronization and discovery, all should be all right.

HOWEVER- Especially if this is happening at 1.2.x, you should look at
system resources (memory!), other applications running, number of
sessions, and the size of the database.  If there's a question about how
to figure the memory you need, please provide some more details about
concurrent operators, database size, all NMS applications, and current
resources.

Sorry if this is too long, but hope it helps.
-------------------------------------------------------------------
Leslie Clark wrote:
>
> The Red process is part of the Management Application Transporter, an add-on 
> to
> Netview that provides support for the Nways Campus Manager product set. It can
> be
> stop/started less traumatically by using the menu-pick under
> 'Tools...Application
> Transporter..Stop/Start' . This process and all of his friends are started 
> when
> the
> map is opened. The list of device-specific applications that are started with
> him can
> be found in the file /usr/lpp/mgtapptran/bin/ed.conf. If they don't go away
> nicely, kill them.
>
> That said, judging by the backlevel condition of your Netview, I would check 
> to
> see if
> your MAT is current, and if the Nways is current. The latest  PTF for Netview
> V4 is
> U451880 (another one coming soon, they say). The latest  MAT is U444149, from
> 1996, however the level that is shipped with Nways Manager for AIX V1.2 is
> mgtapptran.obj 3.0.0.0. The latest PTF for Nways Manager for AIX 1.2 is 
> U456595,
> from 5/98. The Nways Campus Manager product (V3) just went end-of -support
> in May.  With Nways, you always want to be as current as possible. It just 
> keeps
> getting better and better!  There is a chargable upgrade from NCM Lan V3 to
> Nways Manager for AIX 1.2 LAN component.
>
> Cordially,
>
> Leslie Clark
> IBM Global Services - Network & Systems Management - Detroit
> _________________________________________________________________________________________
>
> AIX = 4.2.1.0
> nv6000.base.obj
>           4.1.1.0           COMMIT       COMPLETE     08/29/97     13:50:36
>           4.1.1.0.U443133   COMMIT       COMPLETE     09/15/97     10:04:34
>           4.1.2.0.U449573   COMMIT       COMPLETE     09/15/97     10:04:34
>
> Why is this guy taking on so much time?
>
>  UID     PID  PPID   C    STIME    TTY  TIME CMD
>
> netwrk1 21364 20508  60   Jun 03      - 57:53 /usr/lpp/mgtapptran/bin/Red
> -ft -
> f /usr/lpp/mgtapptran/bin/ed.conf -lc /usr/lpp/mgtapptran/bin/nmvision.ini
>
> netwrk 32638 30260  51   Jun 03      - 2135:19 /usr/lpp/mgtapptran/bin/Red
> -ft
>  -f /usr/lpp/mgtapptran/bin/ed.conf -lc
> /usr/lpp/mgtapptran/bin/nmvision.ini
>
> Why do you think the netwrk ID has more time than the netwrk1 ID?
>
> This server is periodically hanging up and I can't figure out why, but I
> did
> notice this guy out there eating up time.  Our only fix at this time is a
> reboot.
>
> Thanks in advance!

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