RAID Study Group Minutes 3/15/94
John Lohmeyer
jlohmeye at ncr-mpd.FtCollinsCO.NCR.COM
Wed Mar 16 21:03:00 PST 1994
Minutes for RAID working group 03/15/94
Accredited Standards Committee
X3, Information Processing Systems
Doc. No.: X3T10/94-062R0
Date: March 15, 1994
Project:
Ref. Doc.:
Reply to: G. Penokie
To: Membership of X3T10
From: George Penokie and Ralph Weber
Subject: Minutes of RAID Study Group Meeting
March 15, 1994
Agenda
1. Opening Remarks
2. Attendance and Membership
3. Approval of Agenda
4. Report on last RAID Working Group
5. Dual-Controller Issues
6. SCSI Disk Array Model (94-040r2)
7. SCSI-3 Addressing (94-031r2)
8. SDA Commands and Mode Pages (94-042r2)
9. SDA States and Types (94-041r2)
10. Error Handling for SCSI Controllers (94-024r2)
11. Action Items
12. Meeting Schedule
13. Adjournment
Results of Meeting
1. Opening Remarks
George Penokie the RAID Study Group Chair, called the meeting to order at 9:08
am, Tuesday, 15 March 1994. He thanked Qlogic for hosting the meeting.
As is customary, the people attending introduced themselves. A copy of the
attendance list was circulated for attendance and corrections.
It was stated that the meeting had been authorized by X3T10 and would be
conducted under the X3 rules. Ad hoc meetings take no final actions, but
prepare recommendations for approval by the X3T10 task group. The voting
rules for the meeting are those of the parent committee, X3T10. These rules
are: one vote per company; and any participating company member may vote.
The minutes of this meeting will be posted to the SCSI BBS and the SCSI
Reflector and will be included in the next committee mailing.
George stated that this is the 17th meeting of the RAID study group and the
4th joint RAID Advisory Board/RAID study group meeting. The purpose of the
group is to deal with interface issues related to using RAIDs. The study
group will assess the issues and then formulate a strategy for dealing with
them.
2. 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 company to attend and to express their opinion on the subjects being
discussed.
The following people attended the meeting:
RAID Study Group Meeting Attendees
Phone Number -or-
Name S Organization Electronic Mail Address
---------------------- - ------------------------- -------------------------
Mr. Randy Hall Array Technology 303-938-6035
Mr. Thanh Pham AST Research 714-727-6884
Mr. Charles Binford A AT&T/GIS 316-636-8566
Mr. M.W. Jibbe A AT&T/GIS 316-636-8810
Mr. Jeff Brown Clariion-DG 508-898-5006
Mr. Ed Haske CMD Technology 714-454-0800
Mr. Allan Cole CORE International 407-997-6044 x505
Mr. Bill Dallas A Digital Equipment Corp. dallas at wasted.enet.dec.com
Mr. Ralph Weber A Digital Equipment Corp. weber at star.enet.dec.com
Mr. Mitch Rosich EMC Corp. rosich at emc.com
Mr. Howard Grill Formation 609-234-5020
Mr. Bill Hutchison Hewlett Packard Co. hutch at .boi.hp.com
Mr. Andrew Hill HI-Data 617-937-6770
Mr. Paul Boulay Hitachi Computer Products 408-986-9770
Mr. Ken Cummings A IBM Corp. 602-799-2182
Mr. Giles Frazier A IBM Corp. 512-838-1802
Mr. George Penokie P IBM Corp. gop at rchvmp3.vnet.ibm.com
Mr. Kevin Pokorney Intellistor/Fujitsu CPA 303-682-6649
Mr. Dave Fellinger Mega Drive Systems 310-247-0006
Mr. Dick McCormick Peer Protocols 714-476-1016
Mr. Herb Silverman Peer Protocols 714-476-1016
Mr. Tony Fung Storage Concepts 714-852-8511
Mr. Frank Ruthofford Storage Concepts 714-852-8511
Mr. Atul Talati Storage Concepts 714-852-8511
Mr. Robert N. Snively P Sun Microsystems, Inc bob.snively at eng.sun.com
Mr. Joe Molina Technology Forums 507-931-0967
Mr. Gary Watson Trimm Industries trimm at netcom.com
Mr. Peter Dougherty Unisys 714-380-6270
Mr. Duck D Hyatt pool
Mrs. Duck D Hyatt pool
28 people present
2 ducks present
Status Key: P - Principal ** X3T10 status still being defined
A - Alternate
O - Observer
L - Liaison
S,V - Visitor
D - Duck
3. Approval of Agenda
The agenda developed at the meeting was approved.
4. Report on last RAID Working Group
Last RAID Working Group was held in Austin, TX. Several members of the RAID
Advisory Board were present. George stated his intention that the minutes of
the Austin meeting be published on the SCSI Reflector. Questions were raised
about whether that actually occurred. George promised to review his minutes
publication practices.
5. Dual-Controller Issues
George noted that the C-LUI concept may resolve most or all of the dual-
controller issues raised at previous meetings. The C-LUI concept is discussed
in the new SCSI-3 Disk Array Model document (see below). George reported that
Giles Frazier (IBM, Austin) and Doug Hagerman (Digital Equipment Corp.) both
have reviewed the SCSI-3 Disk Array Model from the perspective of dual con-
troller issues. Giles has some small additions that should be added. Doug
believes that C-LUIs cover all his concerns. George is waiting for the dust
to settle.
Giles and George discussed this issues that Giles has raised. The issues
revolve around the master/slave aspects of dual controller implementations.
How does a master exchange its ownership control with a slave? What commands
are used? How do addressing issues (LUN 0) get managed?
At this point, the discussion turned to a picture of a specific single
controller configuration. The various P-LUI and C-LUI components were
identified, described (as to purpose), and named. Then, the single controller
configuration was mirrored into a dual controller configuration. A result of
this was broadening George's experience in dual controller configurations.
By the end of the discussion, George was convinced that most of the dual
controller issues are addressed by the current C-LUI concepts. However,
George felt that additional descriptive text is necessary in the Disk Array
Model to clarify C-LUI application to dual controller configurations.
6. SCSI Disk Array Model (94-040r2)
George presented a revised version of the SCSI-3 Disk Array Model
(X3T10/94-040r2).
The major new concept in the revision of the SCSI-3 Disk Array Model is C-LUI.
A C-LUI is a Component Logical Unit Identifier. A C-LUI represents an entity
that is not identifiable as a SCSI-3 Device Type. To properly contrast with
this, the definition of P-LUI has been changed to make P-LUI represent an
entity that is identifiable as a SCSI-3 Device Type.
George walked through the Model discussing the specific effects of the C-LUI
addition. At several points, corrections were noted to improve the details of
C-LUI definition and usage.
George noted a long-standing convention that he and Doug Hagerman have
discussed. The Model describes redundancy groups as being attached to spares.
Doug prefers attaching spares to redundancy groups. Recently, George realized
the conceptual benefits of using Doug's attachment model. However, George
believes that the benefits of long-term consistence in the Model outweigh the
conceptual simplification provided by reversing what is attached to what.
Questions were raised regarding relationships when A is attached to B and B is
attached to C. Can A be used as a path to gain information about C? C-LUIs
complicate this problem, because C-LUIs can be attached to other C-LUIs.
Using C-LUIs as a mechanism for describing dual controller configurations also
may complicate this issue. George was not certain that a problem really
exists. George promised to work on the issue when/if its status becomes
clearer.
As discussion progressed, the possible need for a new service was uncovered.
The new service would be directed to a P-LUI. The service could be named
"Report P-LUI Attachments". The service would ask the P-LUI to report
everything that is attached to it. Upon further reflection, an equivalent
need was discovered for C-LUIs. George agreed to add the service to the next
revision of the Model.
Questions were raised about what parts of the standard will be optional and
mandatory. George felt that exact definition of optional and mandatory should
be left as one of the last steps in the standardization process. If pressed
to guess, George felt that all the commands and services would be optional.
The addressing and states definitions might be mandatory.
George presented a list of possible C-LUI devices (things). The list was
extracted from a private communication that George received from Doug
Hagerman. George encouraged those attending to send him their lists.
7. SCSI-3 Addressing (94-031r2) Penokie
George presented a revised version of the SCSI-3 Logical Unit Addressing
proposal (X3T10/94-031r2). Concerns were raised about the relationship of
this P-LUI addressing scheme and the vendor-specific schemes already in use.
George noted that the existing vendor-specific schemes will be valid forever,
but he assumes that vendors will gradually convert to the standard
(generalized) scheme.
Other concerns were raised about the reduction of addressable volume set LUNs
|from 64 to 32. George noted that the reduction has value, in the form of the
ability to address V-LUIs and P-LUIs without doing address remaps.
The level of change in 94-031r2 is very small. George declared the document
stable. He said that it will be presented to the X3T10 Plenary for approval
as soon as the SCC project proposal is approved by X3.
8. SDA Commands and Mode Pages (94-042r2)
George presented a revision to the document defining the new commands and mode
pages needed to implement SCSI disk arrays (X3T10/94-042r2). The major source
of changes was the addition of the C-LUI concept.
9. SDA States and Types (94-041r2) Penokie
George presented the latest revision of the document defining states of an SDA
and the types of devices needed within Disk Arrays (X3T10/94-041r2). There
was some discussion of the "base address" wording as opposed to the "P-LUI 0"
wording. In fairly short order, George was convinced to adopt the "base
address" terminology.
George noted several changes resulting from the new C-LUI concept. The entire
list of C-LUI states is new. George solicited proposals for C-LUI states in
addition to those that he has listed.
10. Error Handling for SCSI Controllers (94-024r2) Hagerman (Weber)
Substituting for Doug Hagerman, Ralph Weber presented the latest draft of the
Error Handling for SCSI Controllers document (X3T10/94-024r2). The document
was new to most people present. Ralph describe how the lists of new ASC/ASCQs
on pages 7 through 9 were taken from the vendor-specific ASC/ASCQs of several
products that Digital markets. Ralph also noted the addition of a Disk Array
device type column for the existing ASC/ASCQs.
George said that he wants to scan the lists for duplicates. George described
the normal policy for defining new ASC/ASCQs, which includes a duplicates
check (made by George). No one else had scanned the list for codes of
interest to them. Some real-time scanning occurred during the meeting.
Some people present requested definition of some ASC/ASCQs that represent
x-LUI states (particularly failure-mode states). George said that he will
consider that idea.
It was noted that Jerry Fredin (AT&T/GIS) has a list of ASC/ASCQs based on
contributions from manufacturers at the time the RAB was formed. Although the
list is somewhat moldy now, it still is deemed to be a valuable reference.
Ralph was asked to have Doug contact Jerry.
11. Action Items
a) Penokie Revise the SCSI Disk Array Model (94-040) document based on
input received and toward making it cover dual controller
configurations.
b) Penokie Revise the SDA States and Types (94-041) document.
c) Penokie Revise the SDA Commands and Mode Pages (94-042) document to
account for the "report attachments" service.
d) Hagerman Contact Jerry Fredin (AT&T/GIS) for original manufacture
inputs regarding vendor-specific ASC/ASCQ definitions for
RAID.
e) Hagerman Revise the Error Handling for SCSI Controllers (94-024)
document.
12. Meeting Schedule
The next meeting of the RAID Study Group is planned for April 19, 1994 at
the Hyatt Regency in Bellevue, WA. The meeting is expected to run from
9:00am-5:00pm. This meeting will be a joint RAID Advisory Board Host
Interface Group and X3T10 RAID Study Group meeting.
13. Adjournment
The meeting was adjourned at 3:11 pm. on Tuesday, 15 March, 1994.
More information about the T10
mailing list