[T13] [T13] Zone Type ... the Missing Zone Attribute

Joe Breher Joe.Breher at hgst.com
Fri May 22 08:02:57 PDT 2015


Formatted message: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r1505226_f.htm">HTML-formatted message</a>
Attachment #1: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r1505226_screen_shot_2015-05-22_at_8.58.48_am.png">screen_shot_2015-05-22_at_8.58.48_am.png</a>

Excellent analysis, Ralph.
While I've not done the entire class and object analysis, I rather expect
that Zone Type would be represented a subclass (i.e. specialization) of the
base class of Zone class*. That said, in order for such to be queryable, the
Zone Type needs to exist in some form accessible in implementations that do
not support some form of Run Time Type Identification. Further, such an
attribute would not seem to have potential for conflict with any eventual
class model of a zoned block device.
Accordingly, I see much good with making the Zone Type be another attribute
in the set currently under construction.
*In some cases with the intermediary class Write Pointer Zone class as a
generalization of the ultimate subclass class (Sequential Write Required Zone
class and Sequential Write Preferred Zone class), such Write Pointer Zone
class being in turn a specialization of the Zone class. See sketch below (not
yet to T10 conventions - it's just a sketch done in my preferred syntax-aware
tool).
[cid:4EA27A6F-D9FF-4EDC-A53B-472E4D436239 at localdomain]
The alert will notice that I snuck in an additional attribute of 'write
pointer valid'. I'll try to justify this in a later post.
Joe Breher
Storage Architecture Technologist
Standards Setting Organization
San Jose Research Center
HGST, a Western Digital company
(478) 2-Breher
(478) 227-3437
This e-mail may contain confidential or legally privileged information of
HGST. If you are not the intended recipient, please notify us immediately by
responding to this e-mail and then deleting it from your system.
On May 22, 2015, at 5:48 AM, Ralph Weber
<Ralph.Weber at wdc.com> wrote:
There is a zone attribute missing from the catalog, Zone Type.
Fortunately, no overwhelming justification comes to mind for rewording
anything currently defined by ZBC and ZAC to account for this omission from
the enumeration of zone attributes ... mostly because the Zone Type zone
attribute is already accounted for in the information returned by a REPORT
ZONE command. However, ...
The Zone Type attribute dramatically influences the nature of the zone
attributes that have been a major topic on these reflectors for the past 36
hours, (e.g., zones with aCONVENTIONAL Zone Type do not have a ZC state
machine and thus cannot report any Zone Condition associated with the ZC
state machine).
Unless Joe Breher objects to the addition (probably based on a UML vision of
zones that represents zone type as a class instead of an attribute), the
revised proposals presented to T13 on 5/27 and T10 on 6/1 will add a Zone
Type zone attribute and imbue it with influence over the definitions for all
other zone attributes. {Actually, anybody is free to object on Joe's behalf,
but as the person who has virtually volunteered to propose a UML model for
zones, Joe's vote trumps all others.}
All the best,
.Ralph



More information about the T10 mailing list