Re: [Pgi-wg] Supportive slides: Request routing idea addressed

It provides more information about the 'Request routing' idea addressed by UNICORE/ARC/gLite...
Since when is this a core requirement for dealing with batch submission to a site? Is this not out of scope? You are opening up a can of worms regarding complex workflows etc. Steven

Steven Newhouse wrote:
It provides more information about the 'Request routing' idea addressed by UNICORE/ARC/gLite...
Since when is this a core requirement for dealing with batch submission to a site? Is this not out of scope? You are opening up a can of worms regarding complex workflows etc.
I do not think this is out of scope. In gLite, we already have CREAM which acts as a front-end to multiple batch queues, and the gLite WMS which acts as a front-end to multiple CEs. Submitting a job through "PGI-BES" to the gLite WMS will cause the job to be forwarded to tbe CE better matching the job requirements to be selected for execution, and the job being forwarded there. Even the current, "plain" BES defines the concept of "embedded resources", which in the CREAM-BES prototype we used to let users access specific batch queues. If we consider the "easy" problem of forwarding individual jobs without dependencies (no workflows) I cannot see major reasons not to address this problem. Moreno. -- Moreno Marzolla INFN Sezione di Padova, via Marzolo 8, 35131 PADOVA, Italy EMail: moreno.marzolla@pd.infn.it Phone: +39 049 8277103 WWW : http://www.dsi.unive.it/~marzolla Fax : +39 049 8756233
participants (2)
-
Moreno Marzolla
-
Steven Newhouse