home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.sysv386
- Path: sparky!uunet!infonode!xanth.b11.ingr.com!richard
- From: richard@xanth.b11.ingr.com (Richard Griffiths)
- Subject: Re: Archive Viper 525 and Dell
- Message-ID: <1992Jul29.213646.28863@infonode.ingr.com>
- Sender: usenet@infonode.ingr.com (Usenet Administrator)
- Reply-To: richard@xanth.b11.ingr.com
- Organization: Intergraph Corporation, Huntsville, AL
- References: <1992Jul28.025909.7068@thunder.mcrcim.mcgill.edu> <Bs55tB.6EI@zeus.dialix.oz.au>
- Date: Wed, 29 Jul 1992 21:36:46 GMT
- Lines: 49
-
- In article <Bs55tB.6EI@zeus.dialix.oz.au>, peter@zeus.dialix.oz.au (Peter Wemm) writes:
- |> bruno@thunder.mcrcim.mcgill.edu (Bruno Hall) writes:
- |>
- |> >I am trying to get Dell SVR4 2.1 to talk to an Archive Viper 525.
- |>
- |> >Has anyone managed to do this?
- |>
- |> >At the moment, it'll write a tape file at the beginnig of a tape, and
- |> >then refuse to write another after the first. Also, the light on the
- |> >drive stays on after the initial access -- until the tape is removed.
- |>
- |> Whew! For a while there, I thought I was the only person having
- |> problems with the 525 on Dell. My problems are MUCH worse than yours though,
- |> I cannot even write (there's a blocksize problem somewhere..).
- |> If I write with a blocksize of 20K (the blocksize is irrelevant- i have tried
- |> every different one I could think of), a program I set up to debug
- |> the thing says "write failed: could only write 10240 of 20480 bytes".
- |> Reads are similar: If I read 20K blocks, the read says it read 20K, but in
- |> actual fact only put 10K in the buffer. No matter the blocksize, it
- |> is always out by 50%.
- |> My Viper 2150S Drive works fine.
- |> I have no idea why it does this, but I am getting annoyed enough to start
- |> "hacking". I have though of putting a wrapper around the st01read and
- |> st01write to double and halve the block counts as appropriate - any reason
- |> this wont work?
- |> Also, the drive *floods* the SCSI bus *much* more than the 150 drive..
- |> (Yes, I have tried all sorts of combinations of the jumpers...)
- |>
- |> Anybody else seen this?
- |>
- The Archive 525 I have installed on my Esix 4.04a system works great!
- (apologies to Larry and the Dellevangelists, I just couldn't resist).
- The reason it works great is I modified the driver for the folks at Esix so
- the damn thing would work. Archive screwed up the auto density select.
- The USL scsi tape drivers check the min/max block length returned by the drive
- to decide if they are working with fixed or variable block lengths. If you
- use anything other than a 565mb tape you can only write to it in fixed block
- mode (512 bytes, we're talking hardware level here not cpio/tar). The Archive
- drive always returns the maximum block length the drive is capable of
- regardless of the tape media installed. Stupid. That means the tape driver
- will assume variable block lengths when the media is incapable of it.
- By the way, make sure your rom level on the Archive is at least 7. Prior to
- that auto density did not work at all.
-
-
- --
- Richard A. Griffiths ...uunet!ingr!b11!xanth!richard (UUCP)
- Intergraph Corp. richard@b11.ingr.com (Internet)
- "Something's happening here but you don't know what it is. Do you, Mr. Jones."
-