Minutes of SSC/SMC working group, July 18, 1996 (96-219R0)

[Erich Oetting]

* From the SCSI Reflector, posted by:
* erich_oetting at stortek.com (Erich Oetting)
*
Minutes of SSC/SMC Working Group mtg.     X3T10/96-219r0

Colorado Springs, CO.

Accredited Standards Committee
X3, Information Technology

To:       Membership of X3T10

From:     Erich Oetting

Subject:  Minutes of SSC/SMC Working Group Meeting
          Colorado Springs, CO. July 18, 1996

                          Agenda

1. Opening Remarks

2. Approval of Agenda

3. Attendance and Membership

4. SSC Topics
4.1 - Set Capacity 96-178.
4.2 - Write Protect 96-179.
4.3 - Read Reverse 96.180.
4.4 - Progress Indication. 96-181
4.5 - Serial Issues
4.6 - Sleep/Wakeup for tapes.
4.7 - PLDA proposals for tape
4.8 - More than 256 tape partitions.

5. SMC Topics.
5.1 - World wide names in libraries.
5.2 - Ready/Not Ready ASC/ASCQ for libraries.
5.2 - Publish.
5.3 - Reserve Element (6) and Release Element (6).

6.   Other Topics

7.    Meeting Schedule
8.    Adjournment


                              Results of Meeting

1.    Opening Remarks

Ted Lappin, the SSC Technical Editor, called the meeting to order at 9:00
a.m., Thursday, July 18, 1996.  He thanked John Lohmeyer of Symbios for
arranging and hosting the meeting.

As is customary, the people attending introduced themselves and a copy of
the attendance list was circulated.

The draft agenda was approved.

3.    Attendance and Membership

Attendance at working group meetings does not count toward minimum
attendance requirements for X3T10 membership.  Working group meetings are
open to any person or organization directly and materially affected by
X3T10's scope of work.

The following people attended the meeting:

         Name       Organization        Email Address
 ---------------    -------------       -------------
Mr. Erich Oetting   Storage Technology  erich_oetting at stortek.com
Mr. Edward Lappin   Exabyte Corp.       tedl at exabyte.com
Mr. Arlan Stone     Unisys              arlan.stone at mv.unisys.com
Mr. Paul Entzel     HP                  Paul_entzel at hp.com
Mr. Ralph Weber     Symbios Logic       ralph_weber at symbios.com
Mr. Michael O'Donnel Storage Technology mike.odonnel at stortek.com
Mr. Gary R. Stephens FSI                6363897 at mcimail.com
Mr. Rob Basham      IBM                 robbyb at vnet.ibm.com
Mr. George Penokie  IBM                 GOP at RCHUMP3.VNET.COM
Mr. Charles Monia   Digital             MONIA at shp.dec.com
Mr. William Dallas  Digital             dallas at zk3.dec.com
Mr. Roger Cummings  DPT                 cummings_roger_dpt.com
Mr. Robert Snively  Sun Micro           bob.snively at sun.com

Total of 13 People Present


4. SSC Topics

4.1 - Set Capacity (96-178).

Gary Stevens proposed a SET CAPACITY command to limit tape capacity to a
fraction of normal.  This would be used in applications where access time
is more important than capacity.  The relationship of this command to tape
partitions, tape format commands and the Report Supported Density command
was discussed.  Altering the capacity would clear any existing data and
partitions on tape.  Most existing tape formats will not be able to
support this command as they are not able to remember the capacity
limitation when unloaded and reloaded.  Ted Lappin suggested that this
function be added to the present tape format command.  Discussion then
centered on if Report Supported Density command should report the limited
capacity, or the normal capacity of the tape.  Gary will revise the
proposal to use the FORMAT MEDIA command to reduce the capacity.  Report
Supported Density will return the reduced capacity and a bit will be added
to indicate the capacity has been limited. Gary will bring this to the
September working group meeting for further action.

4.2 - Write Protect (96-179).

Gary  proposed adding several new forms of write protection to tape drives
and cartridges. Mike O'Donnel suggested adding an append only mode.  A
proposal for this will be made in September. Erich suggested changing the
splitting the HARDWARE WRITE PROTECTED asc/ascq to MEDIA WRITE PROTECTED
and LOGICAL UNIT HARDWARE WRITE PROTECTED. Charles Monia suggested some
other minor wording corrections.  Gary will bring the revised proposal,
with ASC/ASCQ and wording changes to the Plenary.

4.3 - Read Reverse (96.180).

Gary proposed adding an option bit to the read reverse command. This would
allow the drive to return data in the same order as a read forward command
instead of reversed.  Gary will revise the document to make it clear that
only supporting the command with the option bit as 1 is permitted.  The
revised proposal will be presented to the Plenary.

4.4 - Progress Indication. (96-181).

Gary proposed adding text to specify how command progress may be reported
after certain immediate commands.  Erich suggested dropping WRITE
FILEMARKS as immediate has a different meaning for this command. How often
a device should update the progress indication was discussed. Gary will
drop the text about SET CAPACITY (command does not exist) from the
proposal.  Other minor wording changes will also be made and the revised
document will be presented to the Plenary.

4.5 - Serial Issues.

Ted talked about the serial issues and that proposed error recovery
mechanisms discussed on the reflector will not work for devices such as
printer.  Charles asked about ways to recovery from transport layer
failures with a tape device.  In most cases disks can simply reissue the
command, but this is not true for most tape devices.  Tape devices can
sometimes recover by using locate commands, but printer devices do not
have this option.

Gary pointed out that fibre channel class 2 solves the problem of knowing
when commands and status are not delivered.  Some details about error
recovery in fibre channel were then discussed.  Roger talked about the
possibility of mixing class 2 and class 3 on a single arbitrated loop.
Bob Snively pointed out that some initiators can mix classes, but others
may not support both classes at the same time.  He also mentioned that the
copy command might be tricky to implement in a mixed class environment.

Arlan pointed out that the working groups assumptions that commands will
not be queued, and delivery will be in order, must be documented. The
group then discussed if this should be in the tape section of PLDA or a
separate white paper.

Bob will write a letter to the reflector suggesting that PLDA allow
retransmission of write data, and that any tape command be completed,
(status returned) before the next command is started.  Bob pointed out
that with these two restrictions, the problems and error recovery methods
used in present parallel SCSI systems would suffice. To use the full
functionality of queued commands, class 2 would be required.

Error recovery was further discussed.  Bob Snively presented scenerios
using queued commands, and showed how over class 3 they may violate the
current tape model.  Gary contended that tagged queuing could actually be
used in class 3. Discussion on this issue continued till it was time to
start the Plenary.

4.6 - Sleep/Wakeup for tapes.

This issue was not discussed due to time limitations.

4.7 - PLDA proposals for tape

This was discussed earlier, see section 4.5

4.8 - More than 256 tape partitions.

How to support more than 256 partitions was not discussed due to time
limitations.

5. SMC Topics.

No SMC topics were discussed due to time limitations, these topics will be
on the September working group agenda.

5.1 - World wide names in libraries.

The use of world wide unique names to identify Data Transfer Elements and
Import/Export Elements. These could replace the function currently
provided by SCSI ID and LUN fields in Data Transfer Elements.  (The SCSI
ID and LUN fields as defined are too small for SCSI-3).

5.2 - Ready/Not Ready ASC/ASCQ for libraries.

Ready/Not Ready ASC/ASCQ codes are not currently available for Media
Changers, but are being used by existing devices.

5.2 - Publish.

Ralph Weber, as SPC document editor, wants SMC published soon due to the
normative referances in SPC to SMC.

5.3 - Reserve Element (6) and Release Element (6).

If SPC makes Reserve (6) and Release (6) optional or obsolete, SMC should
do the same with Reserve Element (6) and Release Element (6).

6.   Other Topics

No other topics were discussed.

7.    Meeting Schedule

The next meeting of the SSC/SMC Working Group will be in Natick,
Massachusetts.  Actual meeting time will be determined by the X3T10
Plenary.

8.    Adjournment

The meeting was adjourned at 1:05 pm on Thursday July 18, 1996.