ADSM-L

Re: ANR8824E Errors on 3494 with 3590E Drives

1999-09-29 17:53:50
Subject: Re: ANR8824E Errors on 3494 with 3590E Drives
From: bbullock <bbullock AT MICRON DOT COM>
Date: Wed, 29 Sep 1999 15:53:50 -0600
        When we first had this problem, I had our IBM CE open a hardware
case with the 3494 folks. They said "it must be your network" and closed the
case. I had our network folks look around and an put some sniffers on the
lines, but they found nothing.

        I have not logged a software call to ADSM support, although if I saw
this problem more often, I would be inclined to do so (have to allocate my
time to the things that keep me awake at night).

The error I am getting is always on this operation, which looks to be a
mount request:

        ANR8824E I/O Error on library TAPELIB2; request some-number for
operation 004C6D31

Just to compare your environment to ours:

        Hardware:       various RS/6000 servers
        AIX OS: 4.2.1 and 4.3.2 , both have logged this error.
        ADSM ver:       3.2.1.40
        Atape ver:      4.4.0.0
        atldd ver:      4.0.1.0
        3590e (FW):     C747

        We have one ADSM server that has not logged any of these errors, and
it's 7 drives are 3490B1A drives but everything else is the same. We will be
upgrading those drives to 3590E1As tomorrow and it will be interesting to
see if I begin to log this error on this server.

Ben


> -----Original Message-----
> From: James SPORER [mailto:james.sporer AT CCMAIL.ADP.WISC DOT EDU]
> Sent: Wednesday, September 29, 1999 2:07 PM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: ANR8824E Errors on 3494 with 3590E Drives
>
>
>      Ben,
>      Hard to believe we are the only two people in ADSM land
> having this
>      problem.  The IBM hardware guys came back and said, the
> first time
>      this happened the tape was already mounted and ADSM
> didn't dismount it
>      before issuing another mount for the same volume.  I
> would guess that
>      could happen occasionaly but it shouldn't be a problem.
> Did you open
>      this as a software problem or a hardware problem?
>      Jim Sporer
>
>
>
> ______________________________ Reply Separator
> _________________________________
> Subject: Re: ANR8824E Errors on 3494 with 3590E Drives
> Author:  "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>  at IPNET
> Date:    9/28/99 1:02 PM
>
>
> We too have 3 3494 libraries with 32 3590E drives. Since we
> installed them
> and upgraded to 3.2.1.40 about 60 days ago, we have seen this
> error 4 times.
> They were scattered across all 3 libraries.
>         We logged a call and IBM analyzed the 3494 log and
> found that the
> library had mounted the tape and replied back to the ADSM
> server that the
> tape was mounted, however, the ADSM server never got the
> message and logged
> an 8824 error.
>         IBM suggested there was some sort of TCP/IP
> communication error that
> caused the problems, but since our libraries are in different
> buildings, on
> different networks and switches, I kinda doubt it.
>
>         Luckily, the error is non fatal. If we make the tape
> available again
> and run the command again, it mounts up fine.
>
> Ben
>
> > -----Original Message-----
> > From: James SPORER [mailto:james.sporer AT CCMAIL.ADP.WISC DOT EDU]
> > Sent: Tuesday, September 28, 1999 11:56 AM
> > To: ADSM-L AT VM.MARIST DOT EDU
> > Subject: ANR8824E Errors on 3494 with 3590E Drives
> >
> >
> >      We recently upgraded two 3590 drives to 3590e drives and
> > added two
> >      more 3590e drives to our 3494 ATL.  Since the change we
> > have gotten a
> >      ANR8824E message on four occasions.  The error causes
> > ADSM to change
> >      the status of the volume it is attempting to mount to
> > unavailable.  We
> >      have reported the problem to our IBM hardware
> > maintenance people and
> >      they are looking at it.  I was wondering if anyone
> else has had a
> >      similar problem?  We are running the AIX version of
> ADSM at level
> >      3.1.2.40.
> >      Jim Sporer
> >      james.sporer AT doit.wisc DOT edu
> >
>