Partial Medium Load for MAM Access

JERMAN,STEVE (HP-Boise,ex1) steve_jerman at am.exch.hp.com
Fri Oct 22 06:15:04 PDT 1999


* From the T10 Reflector (t10 at t10.org), posted by:
* "JERMAN,STEVE (HP-Boise,ex1)" <steve_jerman at am.exch.hp.com>
*
Hi Paul,

my opinions are as follows:

1) I think that the concept of having 'load but not thread' and 'unload but
not eject' are good ideas. They should help support automation use of tape
drives. They are both being done today one way or another.

2) I support the idea that they are only needed for tape devices. I wonder
whether they'll ever invent no-mass tape? Probably at the same time as
teleportation.

3) I would suggest that new commands are not necessary. On a previous DAT
drives we implemented a vendor unique bit in the load/unload command that
accomplished 'unload but not eject'. That seems to work fine. The method was
... Send unload command with 'keep' bit set, drive unloaded tape but did not
eject. Send unload command with keep bit off, drive spat out tape.

4) As for 'load but not thread'. I would acheive this using a mode bit (in
the sequential device page?). If this bit was set then the drive would not
thread tape, but sense data would change for TUR. Host could then send load
command to complete load.

Regards
Steve, HP


-----Original Message-----
From: Paul_A_Suhler at notes.seagate.com
[mailto:Paul_A_Suhler at notes.seagate.com]
Sent: 21 October 1999 10:30
To: Gene_Milligan at notes.seagate.com
Cc: t10 at t10.org; ljesion at us.ibm.com
Subject: Re: Partial Medium Load for MAM Access


* From the T10 Reflector (t10 at t10.org), posted by:
* Paul_A_Suhler at notes.seagate.com
*
Gene raises a good point.  Two things to consider:
   A new command in SPC-2/3 would use up an opcode, but those don't seem to
be
   in short supply.  Or are they?
   Such a command wouldn't help out SMC, unless it were fairly complex.

Off-line, Lee Jesionowski, of IBM, made a number of suggestions:
   This might only be needed for tape devices, since the spin up/down times
for
   other media are relatively short.
   An alternative to modifying CDBs would be to use Mode Select to put the
   device in a mode not to thread and then use the existing Load/Unload.
After
   reading MAMs, do another Mode Select to restore normal functioning.
   A sense key of Not Ready would be more appropriate for the proposed
ASC/ASCQs
   than Unit Attention, since the primary medium is, in fact, not ready.

Comments, please.

Thanks,

Paul



* From the T10 Reflector (t10 at t10.org), posted by:
* Gene_Milligan at notes.seagate.com
*
If it is needed in so many commands sets, why not make it just in SPC-2 or
SPC-3?

Gene


*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org






More information about the T10 mailing list