[ogsa-dmi-wg] DMI Stuff

[Michel Drescher]

Hi all,

once again, please find attached an updated version of the  
specification.

Changes relative to version 55:
- Fixed errors I introduced in version 55
- Redesign the XML Schema according to the specification, following MS  
interop preferences. (trying to)

Please have a good look before our call today!

Cheers,
Michel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ogsa-dmi-spec-56.doc
Type: application/octet-stream
Size: 532480 bytes
Desc: not available
Url : http://www.ogf.org/pipermail/ogsa-dmi-wg/attachments/20080514/c51cdad2/attachment-0001.obj 
-------------- next part --------------


On 13 May 2008, at 22:02, Michel Drescher wrote:

> Hi all,
>
> please find attached a version 55 of the functional specification.  
> As promised I took care of comments 5.9, 7.3 and 7.4.
>
> I will also upload this version to Gridforge (for the earlier  
> version, Gridforge was down).
>
> Cheers,
> Michel
>
> <ogsa-dmi-spec-55.doc>
>
>
> On 13 May 2008, at 18:11, Michel Drescher wrote:
>
>> Steve,
>>
>> taken from Mario's PDF, minus the one(s) I addressed:
>>
>> Issue: 2
>> Reporter: Duane Merrill
>> Comments:
>> 	It's unclear how HTTP/SCP/etc would work.  I could envision the
>> 	destination DEPR having a file://blah protocol so that, if the DMI
>> 	instance was co?located with the sink, you could use http as the
>> 	source.
>> Outcome:
>> 	We should put in file as an example protocol. Also, review
>> 	the text to make clear that things are clear.
>> Status:
>> 	Steven will clarify the exact nature of the
>> 	problem, get in touch with Duane if ncessary
>> 	and sort out the text accordingly.
>>
>> Issue: 4.5
>> Reporter: Rob Schuler and Anne  Chevernak
>> Comments:
>> 	From section 5.2: Support for Stop or Suspend
>> 	Not all transfer protocols will be able to support the stop or
>> 	suspend operations. In such cases, I suppose the DTI state
>> 	becomes failed:unclean or perhaps the service returns an
>> 	exception . I am not clear on that.
>> Outcome:
>> 	We have some state in the core state model that may not
>> 	be supported ? do we have a fault that says we can't move
>> 	into that state? Currently have: IncorrectState fault, failed
>> 	transition fault  ? need in 5.3 a requested state not
>> 	supported fault (RequestedStateNotSupported Fault ).
>> Status:
>> 	Steven to resolve
>>
>> ... and three other minor ones (5.9, 7.3 and 7.4) which I will take  
>> over.
>>
>> Ravi has some issues on his back, too.
>>
>> Cheers,
>> Michel
>>
>> On 13 May 2008, at 17:44, Steven Newhouse wrote:
>>
>>>> Updating the spec:
>>>> Almost done, we "just" need to agree on each other's changes. Only
>>>> some three issues left to address.
>>>
>>> Is there anything else for me to do? Mario assigned me additional  
>>> actions... but I've not had time to look at them yet.
>>>
>>> Steven
>>>
>>>
>>
>