More than an Identifier ...but... less than a Name
joe at q-music.com
Thu Jan 20 16:29:07 PST 2011
* From the T10 Reflector (t10 at t10.org), posted by:
* Joe Breher <joe at q-music.com>
My use of the term 'domain' in my initial post was meant to be taken as
domain in the general sense -- wider than the scope of our defined term of
'SCSI domain'. If a 'snapshot token' (i.e. the handle we use to refer to the
pile of point-in-time logical block copies) crosses SCSI domains, that just
makes the domain (not necessarily SCSI domain) over which names must be
unique and invariant that much larger.
Let me ask this in an attempt to add clarity (for my benefit at least):
This 'snapshotty' thing is assigned a [ name | token | identifier ] by some
entity within a SCSI domain. May this snapshotty thing later need to be
dereferenced in another SCSI domain, using the same [ name | token |
If so, it would seem to me that the name, token, or identifier be created in
some deterministic fashion, such that name conflicts could not occur.
Again, should this be the case, it sounds as if what is needed is a name.
If not however, perhaps 'snapshot identifier' would be appropriate
(substitute the term of your choosing for 'snapshot').
Crawling further out on the limb, to me a 'token' implies something that an
object possesses only long enough to present as a credential to another
object. I don't see it as a unique characteristic of the object to be
identified. Your view may differ.
I still don't see the necessity of a third classifier beyond 'name' and
On Jan 20, 2011, at 4:15 PM, Ralph Weber wrote:
> * From the T10 Reflector (t10 at t10.org), posted by:
> * Ralph Weber <roweber at ieee.org>
> I am not asking for a term for the thing (e.g., snapshot,
> point-in-time-copy, ...).
> I am looking for a term for the 512-byte token whose life
> is bigger than the domain in which it arose (at least in
> those cases were the creator of the token has ports connected
> to multiple domains).
> As of this writing, "token" is still in the pole position.
> All the best,
> On 1/20/2011 4:16 PM, Gerry Houlder wrote:
>> The discussion on this subject during the T10 meeting got better
understanding of the whole "point in time copy of stuff" concept by calling
it a snapshot. Snapshot seems like a perfectly good term.
>> On Thu, Jan 20, 2011 at 2:58 PM, Joe Breher <joe at lingua-data.com
<mailto:joe at lingua-data.com>> wrote:
>> * From the T10 Reflector (t10 at t10.org <mailto:t10 at t10.org>),
>> posted by:
>> * Joe Breher <joe at lingua-data.com <mailto:joe at lingua-data.com>>
>> Hmmm. This could make for an interesting philosophical discussion.
>> In my mind, names are not long-lived so much as they are
>> permanent. The name is a property of an object as long as that
>> object is in existence. Further, names have the characteristic
>> that, for the entire life of any given domain, the name uniquely
>> identifies a given object - whether that object has not yet been
>> instantiated, is currently in existence, or is no longer in
>> existence. That name will never be reused within a given domain.
>> Yes, I acknowledge that several things we refer to as 'names', may
>> not be names per this definition, unless they are fully-qualified
>> names. A true name may contain a time component. A good example of
>> such would be a WWN as per the iSCSI IQN textual format.
>> Identifiers, by contrast, only need be unique within a given
>> domain at an instant in time. They are the 'handle-like-thing' by
>> which we refer to any given object, within a given domain, at a
>> given instant in time.
>> Do we really need an intermediary term? Why is 'identifier' not
>> the proper classifier for the characteristic employed in referring
>> to a 'pile of point-in-time logical block copies'?
>> On Jan 20, 2011, at 12:37 PM, Ralph Weber wrote:
>> > * From the T10 Reflector (t10 at t10.org <mailto:t10 at t10.org>),
>> posted by:
>> > * Ralph Weber <roweber at ieee.org >
>> > *
>> > Here is the perfect kind of grist for the CAP-ites
>> > to chew on.
>> > Today, SCSI has Identifier, which are short-lived numbers
>> > that identify something (e.g., Command IDs, formerly Tags,
>> > identify one command as the Q in an I_T_L_Q nexus).
>> > SCSI also has Names, which are long-lived constructs that
>> > uniquely identify something (e.g., a logical unit name).
>> > What is the right term for something in the middle (neither
>> > short- nor long-lived) that identifies something (e.g.,
>> > the pile of point-in-time logical block copies in a
>> > Copy Offload operation (such as was discussed in Irvine)?
>> > The current proposal says "token".
>> > If there are better ideas, please post them to this reflector.
>> > All the best,
>> > .Ralph
>> > P.S. Whatever gets chosen will be around for a long time.
>> > *
>> > * 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>
> * For T10 Reflector information, send a message with
> * 'info t10' (no quotes) in the message body to 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
More information about the T10