Hi all, I work with computational grids, and I would like to know briefly how the Usage Record Schema is being adopted by the community; who is using it; and in which level the work is currently. Thanks a lot, Glauco
In the NSF's Teragid project, we use the schema to transfer usage data from the resource machines (at 9 different sites) to a central data repository. Different sites have different reporting practices: some sites collect and send atomic data, like CPU hours or wall time; other sites calculate usage based on an internal formula that corresponds to an NSF "service unit". Regardless of how the site measures "usage", the central database can collect a standard usage record for each job and then perform whatever calculations are needed to normalize everything into a single "teragrid unit". This maintains site autonomy while allowing for centralized management. Is this the sort of information you were looking for? thx LM At 10:59 AM 1/18/2005 -0300, Glauco Ludwig wrote:
Hi all,
I work with computational grids, and I would like to know briefly how the Usage Record Schema is being adopted by the community; who is using it; and in which level the work is currently.
Thanks a lot, Glauco
Laura F. McGinnis Project Coordinator: Data & Information Resource Services Pittsburgh Supercomputing Center 4400 Fifth Avenue, # 409D Pittsburgh, PA 15213 email: lfm@psc.edu voice: 412-268-5642 fax: 412-268-5832
The European EGEE project has also adopted the Usage Record schema. The CSAR service (http://www.csar.cfs.ac.uk) generates UR-WG compliant Usage Records for each completed job, and uploads them into a prototype Resource Usage Service (see the RUS-WG working group). The UK's National Grid Service has committed to adopt the Usage Record schema. This is work in progress. Hope this helps, Stephen
-----Original Message----- From: owner-ur-wg@ggf.org [mailto:owner-ur-wg@ggf.org] On Behalf Of Laura McGinnis Sent: 19 January 2005 16:48 To: Glauco Ludwig; ur-wg@gridforum.org Subject: Re: [ur-wg] UR Schema
In the NSF's Teragid project, we use the schema to transfer usage data from the resource machines (at 9 different sites) to a central data repository. Different sites have different reporting practices: some sites collect and send atomic data, like CPU hours or wall time; other sites calculate usage based on an internal formula that corresponds to an NSF "service unit". Regardless of how the site measures "usage", the central database can collect a standard usage record for each job and then perform whatever calculations are needed to normalize everything into a single "teragrid unit". This maintains site autonomy while allowing for centralized management.
Is this the sort of information you were looking for?
thx LM
At 10:59 AM 1/18/2005 -0300, Glauco Ludwig wrote:
Hi all,
I work with computational grids, and I would like to know briefly how the Usage Record Schema is being adopted by the community; who is using it; and in which level the work is currently.
Thanks a lot, Glauco
Laura F. McGinnis Project Coordinator: Data & Information Resource Services Pittsburgh Supercomputing Center 4400 Fifth Avenue, # 409D Pittsburgh, PA 15213 email: lfm@psc.edu voice: 412-268-5642 fax: 412-268-5832
Hi all, DGAS Grid Accounting system currently being integrated in EGEE gLite Candidate Release 1 supports GGF XML UR schema. It treat usage records in such format coming from a computing node. Stores them and can use records to perform aggregate queries of users or resources work and for accounting purposes. Cheers, Andrea Guarise ------------------------------------- Andrea Guarise Istituto Nazionale di Fisica Nucleare Sez. Torino - Via P. Giuria 1 10125 Torino TO ITALY ------------------------------------- On Tue, 18 Jan 2005, Glauco Ludwig wrote:
Hi all,
I work with computational grids, and I would like to know briefly how the Usage Record Schema is being adopted by the community; who is using it; and in which level the work is currently.
Thanks a lot, Glauco
The University of Michigan's MGRID project ingests PBS and Condor usage records into the UR-WG schema, which are then archived and used for accounting and usage statistics. - Beth Kirschner
On Tue, 18 Jan 2005, Glauco Ludwig wrote:
Hi all,
I work with computational grids, and I would like to know briefly how the Usage Record Schema is being adopted by the community; who is using it; and in which level the work is currently.
Thanks a lot, Glauco
Hi Glauco, I have been working on developing a WS-I Resource Usage Service (RUS) based on the GGF rus-wg specification. The RUS stores Usage Records and allows you to query the database of stored records. It has a permission model which ensures privacy of data - you can not see records for which you are not authorised. There is also a simple client (written in Perl) which can insert records into the RUS. This currently runs on the CSAR (www.csar.cfs.ac.uk) service in the UK. There will be a meeting of the RUS WG at the next GGF in South Korea, for which I will produce the interface specification of my RUS implementation. regards John On 18 Jan 2005, at 13:59, Glauco Ludwig wrote:
Hi all, I work with computational grids, and I would like to know briefly how the Usage Record Schema is being adopted by the community; who is using it; and in which level the work is currently. Thanks a lot, Glauco
participants (6)
-
Andrea Guarise -
Beth Kirschner -
Glauco Ludwig -
John Ainsworth -
Laura McGinnis -
Stephen Pickles