Renaming Command Deadlines (14-107)

Ballard, Curtis C (HP Storage) curtis.ballard at hp.com
Fri Oct 17 08:43:05 PDT 2014


Formatted message: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r1410172_f.htm">HTML-formatted message</a>

I really don’t link deadline.  A deadline is a fixed point in time when
something has to be done.  It isn’t a limit on how long something can take.

Webster’s defines deadline as: “a date or time before which something
must be done”

Several people have interpreted this to be creating a method where the system
can provide a timestamp for when the command must be complete and while there
is a definition that says otherwise, a frequent misinterpretation makes
implementation and acceptance more difficult.

I like “Command processing time limit” because that describes exactly
what this is in plain language without having to refer to a definition.  This
is setting a limit on the amount of time the device server can spend
processing the command.  If that is too long “Command time limit” is
further away from “command processing timeout” and shorter.

Paul’s proposal of “Command duration limit” isn’t too bad.

A possible alternative with a nod to the SAM state machine might be
“Command enabled time limit” or “Command enabled limit”.

Curtis Ballard

Hewlett-Packard Company

+1 970 898 3013 / Tel

Curtis.Ballard at hp.com / Email

Fort Collins, CO

From: owner-t10 at t10.org [mailto:owner-t10 at t10.org] On Behalf Of Gerry Houlder

Sent: Friday, October 17, 2014 8:21 AM

To: T10 Reflector

Subject: Re: Renaming Command Deadlines (14-107)

My first choice is to keep the "command deadline" label. We have a
definitions section to describe exactly what we mean.

If that option fails, my second choice is "command time limit". Time limit is
an acknowledged synonym for deadline. Again, if folks need more clarification
than that, read the definition of the term.

On Thu, Oct 16, 2014 at 6:54 PM, Paul Suhler
<Paul.Suhler at hgst.com> wrote:

At Tuesday’s SAM-5 call, I received a recommendation not to use the term
“command deadline time” because that implied an absolute wall clock time,
rather than a duration that begins when the command is received by the device
server.

One suggestion was “command processing time limit.”  This could be
confusing because it’s too close to “command processing timeout,” which
is in SPC-4 and is a totally different concept; see the command timeouts
descriptor in REPORT SUPPORTED OPCODES.  Moreover, “time limit” could
also imply a wall clock time.

Another objection I have is that I’d be replacing three words with four and
this proposal is wordy enough already.

How about:

a)	Command duration limit (Do we *have* to say “processing” – can
it be implicit?)

b)	Command lifetime limit

Other suggestions?

Thanks,

Paul

Paul A. Suhler, PhD

Research Staff Member

HGST Research

paul.suhler at hgst.com

o: 949-476-1180 x275448<tel:949-476-1180%20x275448>

m: 949-241-6443<tel:949-241-6443>

3001 Daimler St.

Santa Ana, CA 92705-5812

www.hgst.com<<a href="http://www.hgst.com/&gt">http://www.hgst.com/&gt;




More information about the T10 mailing list