SAM-2 Revision 11
Ralph Weber
ralphoweber at CompuServe.COM
Fri Sep 10 09:24:19 PDT 1999
* From the T10 Reflector (t10 at t10.org), posted by:
* Ralph Weber <ralphoweber at compuserve.com>
*
George,
Your position on "a Device Server is associated with Zero
or More Logical Units" is a valid interpretation of the model
for implementation purposes. It is not the model and it is
unnecessary for one implementation-based interpretation to
constrict the model.
I am opposed to recasting "the current Target Device as a Bridge
Device with a superclass of Target Device" and recasting "the Target
Manager [a.k.a. Task Manager] as a Bridge Device Server associated
with the Bridge Device" for several reasons. First, the Bridge Device
concept already has meaning as an entity in a parallel SCSI Domain
and reuse of the concept in the model would be confusing. Second,
I don't believe the entire SCSI product engineering community is
ready to embrace the concept of a 'superclass'. As a discussion
on this reflector earlier this year showed, use of the LLP and ULP
concepts are probably unnecessary and confusing in SAM-2. I cannot
support adding yet more computer Ph.D. talk to SAM-2 unless there
is no other way to present the concept.
My use of TST=010b was ill considered. TST=010b is currently reserved
and no one has actually proposed defining it as I suggested. However,
should you wish to see a historical presentation of the definition,
download ftp://ftp.t10.org/t10/drafts/sam/sam-r15.pdf and look in annex
B, specifically on PDF page 87. Remember, this is an early draft of SAM
and this annex was removed before SAM was published.
Thanks.
Ralph...
*
* 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