SBC-2 T10 letter ballot comment resolution

Elliott, Robert (Server Storage) elliott at hp.com
Fri Sep 3 11:12:19 PDT 2004


* From the T10 Reflector (t10 at t10.org), posted by:
* "Elliott, Robert (Server Storage)" <elliott at hp.com>
*
Now available on http://www.t10.org/new.htm are:

sbc2r15a.pdf - Latest working draft, incorporating all comment
resolutions to date
04-274r0.pdf - List of comment resolutions, sorted by author
04-274r0.zip containing 04-274r0.fdf - Adobe Acrobat format comment
resolutions to apply to sbc2r15.pdf (the original revision)

Statistics are:
Total comments: 1449 (459 are editor's notes during incorporation)
 Accept - Done: 1209 (83.44%)
 Accept - Todo: 3 (0.21%)
 Accept (only): 0 (0.00%)
 Reject (only): 195 (13.46%)
         Refer: 28 (1.93%)
  Refer - Vote: 14 (1.93%)
          Left: 0 (100.00%)
  Total worked: 1449 (100.00%)

Please review the resolutions to your and others' comments.

Plan:
* September CAP WG meeting (next week): resolve the outstanding issues,
including holding votes for some items, resulting in the final revision
* November T10 plenary: vote to forward to INCITS for public review


The CAP WG needs to vote on these topics:
* how should the "implicit HEAD OF QUEUE" feature and ORDERED tasks
interact
* require _all_ long LBA commands (e.g. VERIFY (16)) be supported (if
their short LBA peers are supported), or make READ (16) and WRITE (16)
optional
* make READ (6) optional or "deprecated"
* make READ CAPACITY (16) optional
* make REPORT LUNS mandatory
* require that FORMAT UNIT return CHECK CONDITION if RTO_REQ=1 and
FMTPINFO=0
* when RTO=1, allow non-32 byte commands to be used.  They would just
disable checking of the reference tag.  Currently they return CHECK
CONDITION.
* when RTO=0, allow 32-byte commands to be used
* add a unit attention to report that CAPACITY DATA HAS CHANGED whenever
any of the READ CAPACITY
data changes (number of blocks, block size, or various protection
information settings). Is this a "shall" or "should"?
* obsolete support for "volatile medium" such as RAM disks
* obsolete support for removeable media
* obsolete LOCK/UNLOCK CACHE.  If not, decide if locks are I_T nexus
based.
* obsolete PREFETCH
* obsolete most of the Caching mode page bits
* decide if WRITE LONG and READ LONG have implied synchronize cache
behavior

Other outstanding topics:
* block descriptor description is still confusing
* expand the model for memory media
* define "format corrupt"
* is the extra multiplication by x^16 in the CRC checker correct?
* ATO=1 handling in 32-byte commands

-- 
Rob Elliott, elliott at hp.com 
Hewlett-Packard Industry Standard Server Storage Advanced Technology 
https://ecardfile.com/id/RobElliott 
*
* 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