NetNews Usenet Archive 1992 #16/NN_1992_16.iso/spool/comp/sys/next/hardware/1148

Text File | 1992-07-27 | 3.7 KB | 95 lines

Newsgroups: comp.sys.next.hardware
Path: sparky!uunet!mcsun!Germany.EU.net!news.netmbx.de!zrz.tu-berlin.de!math.fu-berlin.de!informatik.tu-muenchen.de!meyergru
From: [e-mail] (Uwe Meyer-Gruhl)
Subject: Re: The DEC DSP3105 3.5" 1G disk keeps crashing.
Keywords: DEC, IBM, HP, 1GByte, 3.5", problems
Followup-To: comp.sys.next.hardware
References:  <[e-mail]>
Originator: [e-mail]
Sender: [e-mail] (USENET Newssystem)
Organization: Technische Universitaet Muenchen, Germany
Date: Mon, 27 Jul 1992 09:39:54 GMT
Message-ID: <[e-mail]>
Lines: 80



In article <[e-mail]>, [e-mail] (Gerben Wierda) writes:
|> So far, we have blown 3 DSP 3105 disks to kingdome come. All of them were
|> built-in in a NeXT (station or cube). All of them stopped working after days
|> or weeks.
|>
|> Another dsp3105 has been working for weeks now in an external enclosue
|> flawlessly.
|>
|> Are there experiences with the reliability of this disk?
|>
|> Do people recognize this?
|> --
|> [e-mail] (Gerben Wierda) || I speak for myself and not for my employer.
|> ([phone]/([phone]  || Support the League for Programming Freedom!
|>     "If you don't know where you are going, any road will take you there."
|>     Lewis Carroll, "Alice in Wonderland".

Hi Gerben and you other folks,


I have tried a DEC DSP 3105 about two weeks ago and have experienced a _big_
problem that kept me from buying it: The drive seemed to work fine, but
every time I copied my old 520 MByte harddisk to the drive, the result was
faulty. After 3 tries, all with different methods (includin dump, gnu-tar
and direct restore from a tape), each of which took about 2 hours to complete,
I eventually found out that the drive returned data I did not write to it.

Yes, this is not a typo. Using gnu-tar with option 'd', I saw several "data
differs" messages.

The drive worked OK in a DEC station, but with only half the write performance,
so I suspect that their kernel does additional verifying.


Now for the good news. After my not-so-good experiences with both the IBM drive
(I posted a while ago) and now with the DEC drive, I think I have found the
right one for me: I recently got an HP C2247. Same capacity as the other ones,
same average access figures, same price (about 3500 DM, but don't quote me
on that).

The drives is just 'plug and play'. The jumpers are documented on top of the
drive. Don't set SDTR (I suspect that this acronym means Synchronous Data Transfer
Request, which won't work, as in my FUJITSU) nor WP (Write Protect).

The drive comes formatted with 512 bytes/sector, and I have not managed to
change this, however the mode sense tells me that I _can_ change the sector size.
Matter-of-fact I have not been able to issue _any_ mode select successfully,
neither with my own programs nor with the new 'scsitools'. Rory Bolt's
'formatter' did not succeed as well.

There is in fact a slight chance that changing the sector size will yield
more space, but there's no guarantee. The IBM drive does change the blocksize
merely logically, so there is no capacity gain. Maybe someone with better
documentation on the HP drive could check this.

The drive performs very well compared to the DEC and the IBM. Its write
performance even beats the read performance. The 'Driveperformance' ratio
is about 1.07, this means it is even faster than a Maxtor 1.4 GByte drive.


hope this helps, Uwe


Uwe Meyer-Gruhl                      "Close female relative wear combat boots"
Lehrstuhl Informatik IX
Technische Universitaet Muenchen     email:[e-mail]
Orleansstr. 34, D-8000 Muenchen 80   tel: ++49 89 48095-209