Glenn Wasson, UVa
Marty Humphrey, UVa
Steven Newhouse, Microsoft
Shahbaz Memon, FZJ
Chris Smith, Platform
Steve McGough, Imperial College London
John Ehrig, OGF
Vesselin Novov, Imperial College London
Paolo Andreetto, INFN
Joel Replogle, OGF
Gabor Robbei, KnowARC
Steve Crumb, OGF
Bill Nitzberg, Altair
SC07 Activity
Flyer : Update awaited from FZJ. Needed by 11:00 Pacific
time TODAY, i.e. NOW!!!
Press Release: Chris and Bill have initiated getting quotes
from their respective organizations. Quotes from MS and OGF in place. Press
release will probably come out Wednesday morning. Discussion will be off list
due to it having embargoed contents.
Demo Updates
·
UVa: FileStaging and ActivityCredential is now
implemented in the compliance tester. Available for interop.
·
Microsoft: FileStaging, ActivityCredential and
AdvancedFilter. Available for interop.
·
FZJ: Endpoint up and running and working with
CREAM BES. Has just got username & password for UVa
·
Platform: Service that is mostly compliant for
username & password. Contact Chris Smith.
·
Imperial College: See wiki page for basic
service details. Different service for file staging. Problems with Platform and
UVa endpoints with the GridSAM client.
·
INFN: Service & FTP server available.
Problems with gLite and the certificate in the compliance tester - unable to
workaround. OK with NorduGrid and GridSAM.
·
KnowARC: Service details up on wiki page. Only
certificate.
·
Altair: Just demoing client side.
BES++ activity at UVa: Deploy BES++ on SGE and PBS. Will
modify the wiki.
Marty will be touring SC07 getting feedback on your
implementation experiences for the document. Please talk to him directly or contribute
thoughts to the list. Good experiences are as important as identifying issues
that needed resolution.
Chris raised an issue as to how to handle the faults that
may be generated on the operations taking a vector of activity identifiers as
input. For HPCBP where the vector length can be constrained to be 1 a normal
SOAP fault can be generated. For BES compliant operations this SOAP fault need
to become (potentially) a vector of SOAP faults. This support is proving
difficult to achieve in gSOAP tooling. In WCF it was also hard to get the
correct classes generated from the XML declaration.