X3T9.2/91-217
Date:        December 9, 1991

To:          X3T9.2 Membership

From:        Lawrence J. Lamers, X3T9.2 Secretary
             George Penokie, Chair Queuing Working Group

Subject:     December 9, 1991 Queuing Working Group Meeting Minutes


George Penokie called the meeting to order at 5:00 p.m. December 9, 1991.  He
thanked Rich Bailey of Vitro Corporation for hosting the meeting.

The final agenda was as follows:

1.     What Should be Said in SCSI-3?
2.     Meeting Schedule

The following people attended the meeting:

             Name              Status          Organization
------------------------------ ------ ------------------------------
Mr. Joe Chen                     A    Cirrus Logic Inc.
Mr. Edward Haske                 O    CMD Technology
Mr. Norman H. Harris             P    Digital Equipment Corp.
Mr. Edward A. Gardner            A    Digital Equipment Corp.
Mr. Paul Hanmann                 P    Emulex Corp.
Mr. Paul R. Nitza                A    Emulex Corp.
Mr. Edward Lappin                P    Exabyte Corp.
Mr. Robert Liu                   P    Fujitsu America, Inc.
Mr. Jeffrey L. Williams          A    Hewlett Packard Co.
Mr. Russell Smith                A    Hewlett Packard Co.
Mr. Giles Frazier                S    IBM Corp.
Mr. Kevin R. Pokorney            A    Intellistor, Inc.
Mr. Geoff Barton                 P    Iomega Corp.
Mr. Lawrence J. Lamers           P    Maxtor Corp.
Mr. Doug Pickford                A    Maxtor Corp.
Mr. John Lohmeyer                P    NCR Corp.
Mr. Steve Kennedy                O    Quantum Corp.
Mr. Gene Milligan                A    Seagate Technology
Mr. Ricardo Dominguez            P    Texas Instruments
Mr. Jeff Stai                    P    Western Digital
Mr. Greg Leonhardt               A    Zenith Data Systems

21 People Present

Status Key:  P     Principal
             A     Alternate
             O     Observer
             S     Special Interest (frequent visitor)
             V     Visitor


Results of Meeting


1.     What Should be Said in SCSI-3?

Should command queuing in SCSI-3 be based on the IPI model of queuing.  The
consensus is yes, the concepts espoused in IPI are the types of queuing that
is needed for SCSI-3.  However, not all variations (e.g., chained commands)
may be supported.  There are also terminology issues because of the different
cultures of SCSI and IPI.  

SCSI has traditionally had the target device manage exception conditions. 
This is not relevant for the IPI model of queuing, because it is a
master-slave architecture.  In as much as possible in SCSI-3 management of
exception conditions should be transparent to the host operating system.  In
order to achieve the desired performance levels it is essential though that
initiators take responsibility for actions and not depend on targets for data
protection.

SCSI has the concept of a linked command that implies a reservation of the
logical unit to an I/O process.  In order to optimize performance in SCSI-3
there is a need to send more than one linked command at a time.  This would
break the implied reservation when linked.  A complete consensus was not
reached on this issue.

IPI has a queuing option of sequential.  This does not directly map to any
SCSI queuing concept, but it is felt that this may be an interesting feature. 
The concept needs some architecture work.

It became clear to the participants that a draft document is needed.  George
Penokie will attempt to draft a document based on his previous work to get us
off the ground in January.


2.     Meeting Schedule

The Queuing Working Group meeting for January is on Monday, January 13, 1992
at 6:00 pm at the Holiday Inn Astrodome Hotel in Houston, TX.

The meeting for February is on Monday, February 17, 1992 at 6:00 pm at the
Doubletree Hotel in Austin, TX.