ADSM-L

Re: DB2 on Solaris 7

2003-12-18 13:15:14
Subject: Re: DB2 on Solaris 7
From: John Monahan <JMonahan AT COMPURES DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Thu, 18 Dec 2003 12:23:00 -0600
"ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU> wrote on 12/18/2003
06:48:14 AM:

> >... ANR2579E Schedule SAP_ONLINE in domain TDP_SERVERS for node
> SAPQA-TDP failed (return code 1).
>
> John - This probably reflects a POSTSchedule or PRESchedule command
> failure, where
>        return code 1 typically accompanies "Command not found".
> Check that out.

The schedule runs a shell script (action=command) and the shell script does
its own error checking and exits with a return code of 1 if there is a
problem, that is where that return code is coming from.

>
> If your Solaris sessions remain problematic, and neither the client
> nor server end
> reflects an action by its end in causing session severance, then it
> would seem that
> something in between is causing it.  Check your Solaris syslog for
> error indications,
> which may have affected things other than the TSM session, too.
> We've seen a good
> number of Solaris-related postings where session problems were
> caused by ethernet
> autonegotiation being in effect.  Check the network configuration of
> that Sun box
> as compared with other Suns not exhibiting the problem, and compare
> commonality of
> subnet/switch connection.  You could also conduct beefy, manual TSM
> sessions, such
> as large Backups, to try to make the problem occur under observation.
>
>   Richard Sims, BU

The syslog is pretty much useless, nothing but sendmail entries.  The
backup runs fine if I run it manually, it also works if I schedule it to
run on its own with no other backups running at the same time.  The
configuration of this box is identical to the other 2 Suns, in the same
switch also.  I thought network problems, but I have run the backup
manually at least 20 times without a single problem.

I have it scheduled on its own for now as a work around.  I think the
problem is that the api/DB2 combination is not waiting properly when the
session goes to media wait for longer than 5 minutes or so.

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