ZBC - FINISH ZONE on a closed zone (late LB comment)

Black, David david.black at emc.com
Tue Mar 3 18:49:20 PST 2015


Formatted message: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r150303c_f.htm">HTML-formatted message</a>

> Though it currently elides the detail of transitioning through Implicit
Open to get there.
> We had thought that the intermediate state of Implicit Open would be
unobservable from the interface,
If that intermediate state has to be unobservable, there would need to be a
"shall" to that effect, and perhaps a direct state machine transition from
Closed to Full - keep in mind that the I in SCSI stands for Interface ;-).
It'd be better to document the intermediate state - if finishing a zone takes
a while, a REPORT ZONES command may see the intermediate implicit open state.
 Also, the implicit open could cause a zone management operation that closes
another zone, and that side effect would be observable.
Thanks,
--David
From: Joe Breher [mailto:Joe.Breher at hgst.com]
Sent: Tuesday, March 03, 2015 8:15 PM
To: Curtis Stevens
Cc: Black, David; T10 Reflector
Subject: Re: ZBC - FINISH ZONE on a closed zone (late LB comment)
Indeed, the FINISH ZONE description states a shall:
The finish zone operation shall cause the zone condition to be set to the
ZONE FULL condition for the specified zone.
... with explicit disclaimers Read Only and Offline.
Though it currently elides the detail of transitioning through Implicit Open
to get there.
We had thought that the intermediate state of Implicit Open would be
unobservable from the interface, and therefore not need to appear in the
command description, but we might concur with some rewording.
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 Mar 3, 2015, at 4:29 PM, Curtis Stevens
<curtis.stevens at wdc.com> wrote:
* From the T10 Reflector (t10 at t10.org<mailto:t10 at t10.org>), posted by:
* Curtis Stevens <curtis.stevens at wdc.com>
*
In the second sentence of the Finish Zone command, there is a reference to
performing a zone management operation.  There are also a variety of letter
ballot comments asking for this to be reworded.  I suspect that this is where
the clarification you are looking for should reside.
-------------------------------------------------
Curtis E. Stevens
Director, Standards & Features Technology
3355 Michelson Dr. #100
Office: 1-1041
Irvine, Ca. 92612
Phone: 949-672-7933
Cell: 949-307-5050
E-Mail: Curtis.Stevens at WDC.com
Remember, you may only be blamed for something if you are actually doing
something.
-----Original Message-----
From: Black, David [mailto:david.black at emc.com]
Sent: Tuesday, March 03, 2015 3:24 PM
To: Curtis Stevens; T10 Reflector
Subject: RE: ZBC - FINISH ZONE on a closed zone (late LB comment)
I agree with not putting state transition information in the command clause.
It should suffice to say that if the zone is closed, the FINISH ZONE command
implicitly opens the zone before performing the finish zone operation.
Thanks,
--David
-----Original Message-----
From: Curtis Stevens [mailto:curtis.stevens at wdc.com]
Sent: Tuesday, March 03, 2015 6:18 PM
To: Black, David; T10 Reflector
Subject: RE: ZBC - FINISH ZONE on a closed zone (late LB comment)
I will enter the comment.
I would like to use this thread to work some of the issue...  We do
not put state transition information into the commands.  We put these
in the state machine, as it currently is.
-------------------------------------------------
Curtis E. Stevens
Director, Standards & Features Technology
3355 Michelson Dr. #100
Office: 1-1041
Irvine, Ca. 92612
Phone: 949-672-7933
Cell: 949-307-5050
E-Mail: Curtis.Stevens at WDC.com
Remember, you may only be blamed for something if you are actually
doing something.
-----Original Message-----
From: Black, David [mailto:david.black at emc.com]
Sent: Tuesday, March 03, 2015 3:14 PM
To: Curtis Stevens; T10 Reflector
Subject: RE: ZBC - FINISH ZONE on a closed zone (late LB comment)
That would make the specification of the actions of the FINISH ZONE
command incomplete, as there is no mention of implicit zone open in
either the specification of the FINISH ZONE command or the definition
of the finish zone operation.
Please enter the late letter ballot comment.  It is ok to resolve it
by specifying that when applied to a closed zone, the FINISH ZONE
command or the finish zone operation implicitly opens the zone.  I
would prefer to see this specified as part of the FINISH ZONE command
in order to avoid expanding the definition of the finish zone operation.
Thanks,
--David
-----Original Message-----
From: Curtis Stevens [mailto:curtis.stevens at wdc.com]
Sent: Tuesday, March 03, 2015 5:31 PM
To: Black, David; T10 Reflector
Subject: RE: ZBC - FINISH ZONE on a closed zone (late LB comment)
David, if you check the zone condition statement machine ZC4: Closed to ZC2:
Implicit open (4.3.3.2.5.3), you will see that there is a transition
for FINISH ZONE in the closed state.  The transition to implicit
open is important...  FINISH ZONE may cause writing to occur in
these devices.	As a result, open zone resources are required to
perform a finish
zone operation.
It is possible that the FINISH ZONE command fails on a drive with
too many explicitly open zones.
-------------------------------------------------
Curtis E. Stevens
Director, Standards & Features Technology
3355 Michelson Dr. #100
Office: 1-1041
Irvine, Ca. 92612
Phone: 949-672-7933
Cell: 949-307-5050
E-Mail: Curtis.Stevens at WDC.com
Remember, you may only be blamed for something if you are actually
doing something.
-----Original Message-----
From: owner-t10 at t10.org<mailto:owner-t10 at t10.org> [mailto:owner-t10 at t10.org]
On Behalf Of
Black, David
Sent: Tuesday, March 03, 2015 11:17 AM
To: T10 Reflector
Subject: ZBC - FINISH ZONE on a closed zone (late LB comment)
* From the T10 Reflector (t10 at t10.org<mailto:t10 at t10.org>), posted by:
* "Black, David" <david.black at emc.com>
*
In a review with some of our engineers, we found a missing case in
the FINISH ZONE command that looks like it'll involve state machine
changes.  Please consider the following to be a late letter ballot
comment from EMC (caveat - I haven't checked this for overlap with
other
letter ballot comments):
-------------
The effect of a FINISH ZONE command effect on a closed zone is unspecified.
This should not be an error, and it should cause a closed zone to
transition to full (even though no such transition is in the current
state machine) because ...
... if a zone is implicitly opened, the application client expects
to be able to issue a FINISH ZONE command to make that zone full.
If an implicitly opened zone is implicitly closed by a zone
management operation, the application client is unaware of that zone
state change, and hence still expects to be able to issue a FINISH
ZONE command.  The result should not be an error (e.g., ILLEGAL
REQUEST is definitely the wrong answer), and the application client
still expects the
result to be a full zone.
-------------
Thanks,
--David
----------------------------------------------------
David L. Black, Distinguished Engineer EMC Corporation, 176 South
St., Hopkinton, MA  01748
+1 (508) 293-7953	      FAX: +1 (508) 293-7786
david.black at emc.com        Mobile: +1 (978)
394-7754
----------------------------------------------------
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to
majordomo at t10.org<mailto:majordomo at t10.org>
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to
majordomo at t10.org<mailto:majordomo at t10.org>



More information about the T10 mailing list