ADSM-L

Re: DLT7000 drives with ADSM

1997-07-28 09:48:08
Subject: Re: DLT7000 drives with ADSM
From: Steven P Roder <tkssteve AT ACSU.BUFFALO DOT EDU>
Date: Mon, 28 Jul 1997 09:48:08 -0400
On Fri, 25 Jul 1997, Pittson, Timothy ,HiServ/NA wrote:
>
> I'm looking at throwing some additional hardware at our ADSM server.
> We're currently using a 3494 with 2 3590 tape drives and have been very
> happy with the performance and reliability.  However, my management
> won't go for another 3494 frame with additional 3590's because of the
> cost.  This leaves me with DLT and 3570 technology as options.  Anybody
> out there in ADSM land using DLT7000 drives  ??  If so,   have you been
> happy with the reliability, performance, etc.  Any gotchas I need to
> watch out for ??

I mulled responding to this over the weekend, and the continued problems
have made me decide to say something here.  I received two BreeceHill Q7's
with DLT7000 drives in them back in May, and they are still not working
reliably.  First, I had problems with PTI's differential scsi card, which
prevented me from writing to the drives (scsi adapter errors).  We got rid
of the PTI boards, replaceing them with differential cards from Sun.  We
then proceeded to get read errors auditing the volumes created.
BreeceHill sent me out new firmware, a replacement drive, and a
replacement power supply, and the errors persist.  I have seen similar
problems accross both Q7's.  The problem seems to crop up as a
calibration error, which then causing subsequent reads to fail.
BreeceHill has now involed Quantum in this, so maybe it will eventually
get fixed.  I think there are still bugs in the firmware and drives
relating to calibration and head positioning.  I do not see this being
resolved anytime soon.  If I were to make a recommendation to management
such as you, with my experiences so far with the DLT7000's, I would tell
them to bite the bullet, and buy more 3494/3590 technology.  In my
opinion, the DLT7000's were rushed to market, and are unusable in their
current state.  They should still be in the lab, not in customer shops.

Also, when these errors first came up, I opened a problem with IBM's ADSM
support, thinking that the problem may be with their device drivers.
After getting some traces, IBM said it was hardware, not software.  I had
an extremely difficult time convincing the hardware vendors that they had
a problem.  Throughout this whole ordeal, I have kept IBM in the loop, and
I would like to say, publicly, that the ADSM Level 2 support team has done
an outstanding job with this problem.  I do not know how anyone could bash
their support services, as was done on this list.  My experiences with
ADSM support have been nothing short of outstanding, partciularly Larry
and Andy.

If anyone else is using DLT7000's, and can say that they work, I would
love to hear about it!  If anyone else is using DLT7000's with similar
problems to mine, I would love to hear about that too!

Steve (unVMix Systems Programmer/Dude) Roder
(tkssteve AT ubvm.cc.buffalo DOT edu | tkssteve AT acsu.buffalo DOT edu | 
(716)645-3564 ,
   | http://ubvm.cc.buffalo.edu/~tkssteve)
<Prev in Thread] Current Thread [Next in Thread>