UWS pattern makes scratch VOSpace redundant

Guy Rixon gtr at ast.cam.ac.uk
Sun Oct 14 04:52:50 PDT 2007


The results URI for a UWS doesn't have to be in the same scheme as the
controls for the URI: its doesn't have to be a HTTP URL. The service provider
can put a vos:// URI for the results.

Cheers,
Guy

On Fri, 12 Oct 2007, Paul Harrison wrote:

> Hi,
>
> I have been thinking about one of the use cases that we been
> advocating for VOSpace- that of using it for temporary storage of
> results between steps in a workflow. This has been a principal use of
> MySpace within Astrogrid for instance. The UWS pattern allows for
> each service to be the  temporary storage for the data that is to be
> the input to the next step in the workflow i.e. the second
> application can read data from the (jobs)/(jobid)/results uri of the
> first step. This is both easier to implement and more efficient in
> data transport terms than moving the data into a scratch VOSpace. UWS
> provides all of the data management facilities needed for this simple
> scenario.
>
> The above conclusion holds for a set of distributed UWS services an a
> centralized VOSpace - if services are co-located then the conclusions
> become more complex. Imagine that there is a UWS data processing
> service and a co-located VOSpace service, where the two services
> actually share the same backend storage - i.e. in VOSpace terms the
> file: protocol could be used by the UWS to retrieve the data. In this
> case there could be some benefit to having a co-located VOSpace
> because  the data could be retrieved and put to the VOSpace very
> efficiently and it allows for long term storage, so that the combined
> UWS/VOSpace service could gradually accumulate raw and processed data
> products that perhaps with the addition of a colocated Querying
> service could provide a valuable dataset.
>
> Cheers,
> 	Paul.
>
> .
> Dr. Paul Harrison
> JBCA, Manchester University
> http://www.manchester.ac.uk/jodrellbank
>
>
>

Guy Rixon 				        gtr at ast.cam.ac.uk
Institute of Astronomy   	                Tel: +44-1223-337542
Madingley Road, Cambridge, UK, CB3 0HA		Fax: +44-1223-337523



More information about the grid mailing list