ADSM-L

Re: ADSM and 3570 MP

1999-04-14 07:30:10
Subject: Re: ADSM and 3570 MP
From: Wassim Hossari <whossari AT KFSHHUB.KFSHRC.EDU DOT SA>
Date: Wed, 14 Apr 1999 14:30:10 +0300
Hi Rui,

Did you talk IBM into replacing the 3570 with another 3570 or another tape
library model? because it seems that we are heading that way.  The IBM CEs had
tried replacing almost every part in the 3570 that you could imagine.  They even
replaced some parts like the Picker Assembly and the Drive Analog Card more than
once.


Regards,
Wassim Hossari

----------------------------------------------------------------------------------------
Rui Malheiro wrote:
Rui Malheiro wrote:

> Replying to: Wassim Hossari <whossari AT KFSHHUB.KFSHRC.EDU DOT SA> (Wed, 14 
> Apr
> 1999 12:24:16 +0300)
>
> >Hi everyone:
> >
> >We are  using ADSM 3.1.1.0 on IBM R50 machine running AIX 4.2.1 to
> >backup our network servers. Everything was going  smooth and well for a
> >long time until we decided to replace the 8mm tape drive with a Magstar
> >MP 3570 Tape Subsystem.  For the last four months after that
> >replacement, IBM CEs had been trying to fix all kind of errors that are
> >generated on daily bases by the 3570.  The advise that I am seeking from
> >any user that have experience with ADSM and the above mentioned tape
> >subsystem is the recommended block size setting for the 3570 tape
> >drives.  Ours now is set to "0" which means variable length block size.
> >Would that be O.K. or should it be set to some other value such as 512
> >or 1024.  I appreciate it if anybody would like to share any experience
> >with regard to dealing with errors generated by the 3570 MP as a tape
> >subsystem to ADSM.
>
> I've had some problemas with a Magstar 3570 last year. The problem was that
> writes to tape would fail with undetermide errors:
>
> 07/14/1998 08:59:39  ANR8302E I/O error on drive RMT3 (/dev/rmt3) (OP=READ,
> CC=0, KEY=03, ASC=11, ASCQ=01, Description=An undetermined error has
> occurred).  Refer to Appendix B in the 'Messages' manual for recommended
> action.
>
> The CE tried everything, from microdode refresh to parts replacement (nearly
> all parts were replaced). In the end we talked IBM into replacing the 3570
> library itself by a new unit, and since then the problem has stoped.
>
> --
> Rui Malheiro,
> 6 Mil - Tecnologias de Informacao
> URL: <http://www.6mil.pt/>
<Prev in Thread] Current Thread [Next in Thread>