From jhs@jhs.muc.de  Mon May 18 13:00:28 1998
Received: from colin.muc.de (root@colin.muc.de [193.174.4.1])
          by hub.freebsd.org (8.8.8/8.8.8) with SMTP id NAA17009;
          Mon, 18 May 1998 13:00:18 -0700 (PDT)
          (envelope-from jhs@jhs.muc.de)
Received: from jhs.muc.de ([193.174.4.84]) by colin.muc.de with SMTP id <140575-2>; Mon, 18 May 1998 19:06:28 +0200
Received: (from jhs@localhost)
	by jhs.muc.de (8.8.7/8.8.7) id QAA00732;
	Mon, 18 May 1998 16:32:32 +0200 (MET DST)
	(envelope-from jhs)
Message-Id: <199805181432.QAA00732@jhs.muc.de>
Date: Mon, 18 May 1998 16:32:32 +0200
From: jhs@FreeBSD.ORG
Reply-To: jhs@FreeBSD.ORG
To: FreeBSD-gnats-submit@freebsd.org
Subject: RELENG_2_2/doc/handbook/handbook130.html PC Hardware compatibility
X-Send-Pr-Version: 3.2

>Number:         6681
>Category:       kern
>Synopsis:       Tapes written on 2.2.8 not readable on 3.2
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    mjacob
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          change-request
>Submitter-Id:   current-users
>Arrival-Date:   Mon May 18 13:00:01 PDT 1998
>Closed-Date:    Sun Jun 6 16:28:57 PDT 1999
>Last-Modified:  Sun Jun  6 16:29:56 PDT 1999
>Originator:     Julian Stacey <jhs@freebsd.org>
>Release:        FreeBSD 2.2 i386
>Organization:
FreeBSD
>Environment:

	I did a
        cvs export -r RELENG_2_2 doc ; make
	and read
	RELENG_2_2/doc/handbook/handbook130.html

>Description:

	I suggest you append some of this info:

        (ncr0:5:0): "DEC TLZ06     (C)DEC 0389" type 1 removable SCSI 2
        st0(ncr0:5:0): Sequential-Access
        st0(ncr0:5:0): 5.0 MB/s (200 ns, offset 8)
        density code 0x0,  drive empty

        DAT drive, does 2 Gig with 90M tape, or more with compression on.
        SIngle slot height.

        ----------

        (ncr0:6:0): "TANDBERG  TDC 3800 =04Y" type 1 removable SCSI 2
        st1(ncr0:6:0): Sequential-Access
        st1(ncr0:6:0): asynchronous.
        density code 0x0,  drive empty

        A 525 M QIC  drive,
        reads 45 & 60M & 150M & 525M,
        writes 150 & 525,
	does variable block sizing by default
		(mt blocksize 512 to suppress that).
        Single slot height.

        ------------

        (aha0:3:0): "  " type 1 removable SCSI 1
        st0(aha0:3:0): Sequential-Access density code 0x5,  drive empty

        A Cipher QIC-24 SCSI, does 60 Meg
        Has a twist lever like a 5" floppy, & a pull out drawer.
        Double height ie 2 slots.

        Reported by: Julian H. Stacey  <jhs@freebsd.org>
Date: Tue, 1 Jun 1999 09:52:07 +0100
From: nclayton@lehman.com
To: scsi@FreeBSD.ORG
Subject: Tapes written on 2.2.8 not readable on 3.2 (was Re: PR docs/6681)

Hi folks,

I'm not on the SCSI list myself, I'm forwarding this on behalf of Julian
Stacey.

I've been closing doc PRs, including docs/6681, and contacted Julian to see
if his PR was still relevant.

In what follows, "> >" is my message to Julian, ">" is his reply.  If any
of you SCSI gurus has the time could you take a look at this?

Cheers,

N

On Tue, Jun 01, 1999 at 12:10:22AM +0200, Julian Stacey wrote:
> > It's this (from the PR);
> > 
> >         I suggest you append some of this info:
> > 
> >           (ncr0:5:0): "DEC TLZ06     (C)DEC 0389" type 1 removable SCSI 2
> >           st0(ncr0:5:0): Sequential-Access
> >           st0(ncr0:5:0): 5.0 MB/s (200 ns, offset 8)
> >           density code 0x0,  drive empty
> > 
> >           DAT drive, does 2 Gig with 90M tape, or more with compression on.
> >           SIngle slot height.
>
> Works on 2.2.8 
> 
> >           (ncr0:6:0): "TANDBERG  TDC 3800 =04Y" type 1 removable SCSI 2
> >           st1(ncr0:6:0): Sequential-Access
> >           st1(ncr0:6:0): asynchronous.
> >           density code 0x0,  drive empty
> > 
> >           A 525 M QIC  drive,
> >           reads 45 & 60M & 150M & 525M,
> >           writes 150 & 525,
> >           does variable block sizing by default
> >                   (mt blocksize 512 to suppress that).
> >           Single slot height.
>
> Works on 2.2.8
> 3.2-RELEASE Seems to move the drive, but couldn't read tapes written 
> on 2.2.8
> 
> >           (aha0:3:0): "  " type 1 removable SCSI 1
> >           st0(aha0:3:0): Sequential-Access density code 0x5,  drive empty
> > 
> >           A Cipher QIC-24 SCSI, does 60 Meg
> >           Has a twist lever like a 5" floppy, & a pull out drawer.
> >           Double height ie 2 slots.
>
> Works on 2.2.8 
> 
> > I know that the switch to CAM broke some stuff, but I'm not sure how
> > extensive the breakage is -- if these aren't affected, and work with 
> > -stable/-current then I'll update the Handbook.
> 
> I have 2 TANDBERG TDC 3800 drives, one on 
>
>       3.2-RELEASE
>       sa0: <TANDBERG TDC 3800 =04Y> Removable Sequential Access SCSI-2 device 
>       sa0: 4.032MB/s transfers (4.032MHz, offset 7)
>
> one on
>
>       2.2.8-RELEASE
>       (ncr0:6:0): "TANDBERG  TDC 3800 =04Y" type 1 removable SCSI 2
>       st1(ncr0:6:0): Sequential-Access 
>       st1(ncr0:6:0): asynchronous.
>       density code 0x0,  drive empty
> 
> Seems to me on brief experiment,
>  3.2 can write data that 2.2.8 can read,
> but 2.2.8 tapes are not readable on 3.2 system
> (a disaster for anyone relying on backup tapes (not me thankfully)).
> 
> Anyway, that was just a quick check,
> maybe it's to do with the blocksize 512 stuff,
> but looks like this is a problem.
> 
> As I have 2 drives, 1 on 2.2.8 & 1 on 3.2, I may be uniquely useful
> for test purposes for scsi team, feel free to pass my name along.
> I am not worried about installing current on the 3.2 Tandberg box,
> except that can't spare too much time for it.
> I am not prepared to destabilise my other main 2.2.8 Tandberg box though :-)
> 
> In light of this nasty suprise I have cancelled plan to upgrade my 2.2.8
> main  system (that has the DEC TLZ06 DAT) until this problem is solved.
> Once it's solved for Tandberg, I can try.
> 
> The system with the Cipher QIC-24 will be upgraded to 3.2 in about 2 weeks.
> & I can let you know about that.
> 
> What was intended as a docu. upgrade has become a warning of peril :-)
> Thanks, I'm glad to know, let me know if I can help,
> feel free to cut/paste forward to scsi@ (not on that list myself).

>How-To-Repeat:

	-

>Fix:
	


>Release-Note:
>Audit-Trail:
Responsible-Changed-From-To: freebsd-doc->mjacob 
Responsible-Changed-By: mjacob 
Responsible-Changed-When: Thu Jun 3 17:22:26 PDT 1999 
Responsible-Changed-Why:  
I own the tape driver 
State-Changed-From-To: open->feedback 
State-Changed-By: mjacob 
State-Changed-When: Sun Jun 6 15:03:33 PDT 1999 
State-Changed-Why:  
Waiting for submitter feedback 
State-Changed-From-To: feedback->closed 
State-Changed-By: mjacob 
State-Changed-When: Sun Jun 6 16:28:57 PDT 1999 
State-Changed-Why:  
According to jhs, Joerg's suggestion worked. I've updated the sa(4) man 
page to make a similar suggestion. 
>Unformatted:
 >From: Matthew Jacob <mjacob@feral.com>
 >To: jhs@freebsd.org
 >Subject: Re: Tapes written on 2.2.8 not readable on 3.2 (was Re: PR docs/6681)
 >    (fwd)
 >
 >
 >
 >Did you try Joerg's suggestion?
 >
 >
 >---------- Forwarded message ----------
 >Date: Wed, 2 Jun 1999 08:05:06 +0200
 >From: J Wunsch <j@uriah.heep.sax.de>
 >Reply-To: Joerg Wunsch <joerg_wunsch@uriah.heep.sax.de>
 >To: nclayton@lehman.com
 >Cc: scsi@FreeBSD.ORG
 >Subject: Re: Tapes written on 2.2.8 not readable on 3.2 (was Re: PR docs/6681)
 >
 >As nclayton@lehman.com wrote:
 >
 >> > Seems to me on brief experiment,
 >> >  3.2 can write data that 2.2.8 can read,
 >> > but 2.2.8 tapes are not readable on 3.2 system
 >> > (a disaster for anyone relying on backup tapes (not me thankfully)).
 >
 >I think that's the well-known problem (*) that the defaults of 3.x and
 >-current differ for some tape drives, for whatever reason.  In
 >particular, the current driver doesn't trust the user ;-) to default
 >to variable mode for QIC > 150, as the pre-CAM tape driver did.  Thus,
 >tapes written in variable mode previously aren't readable by default.
 >
 >However, the world is not that bad, the old tapes are readable without
 >problems if you use `mt blocksize variable' before reading them.  At
 >least, i haven't seen a tape so far i could not read that had been
 >written with the old system.
 >
 >(*) Examining the archive of this list for the subject `tape' should
 >reveal you a number of threads discussing it.
 >
 >-- 
 >
 >
