Hi Andrea, John, Jules,... Am 23.03.2010 16:58, schrieb Jules Wolfrat:
John, Andrea,
-----Original Message----- From: ur-wg-bounces@ogf.org [mailto:ur-wg-bounces@ogf.org] On Behalf Of john.gordon@stfc.ac.uk Sent: 23 March 2010 16:42 To: andrea.cristofori@cnaf.infn.it; ur-wg@ogf.org Subject: Re: [UR-WG] OGF28 Schedule
Andrea, at OGF28 someone suggested that we had a choice of either fitting the storage record into the existing UR or defining a new standard. I do not think that this is the case. The existing UR UsageRecord element has a UsageRecordType type of which JobUsageRecord is a defined type. I think we need to define a new type StorageUsageRecord which can contain some of the existing Global attributes defined for the JobUsageRecord.
So, apparently the choice already was made, with good arguments!
I am not sure whether a choice has already been made, but I like to second John's and Jules' statements and arguments here. As far as I can see, the general concept of the usage of compute resources is orthogonal to the general concept of the usage of storage resources. This is what we discovered and discussed step by step during the meeting last week - I think. Of course we will have go through the details and should analyze some variety of realistic use cases before we should make the final design decisions. However, currently I have no doubt that we might end up with a resource usage model that will require for different schemas, one for the description of the consuption of compute resources where the attributes startTime and endTime make sense and another for the current consumption storage resources which could be represented by long-time archives as well as by short.time storages. Regards, Johannes
Regards,
Jules
We could also define a FileUsageRecord to cover the case you were defining Andrea.
John
-----Original Message----- From: ur-wg-bounces@ogf.org [mailto:ur-wg-bounces@ogf.org] On
Behalf Of
Andrea Cristofori Sent: 23 March 2010 14:49 To: ur-wg@ogf.org Subject: Re: [UR-WG] OGF28 Schedule
Dear all,
While waiting for the minute I'll suggest we can already start to discuss one basic topic that came up during the meeting before we
move
to more specific things.
- Since the beginning I was thinking just to add new properties
to the
existing UR definition. This would allow to choose those that are needed in each situation and reuse those that are common. In case of new versions of the UR definition if we update a property then is not needed to update all the definition if this is common. Not all of us
agreed
with that and it was proposed to create a different usage record
for
the Storage. What do you think?
- What would you prefer to do in case we discuss other property
not
directly connected with storage accounting? Should we include everything in the discussion now or concentrate only on this topic and try
to
finalize an update before moving to other things?
Cheers Andrea
On 03/18/2010 05:19 PM, Andrea Cristofori wrote:
P.S.: here is the link
http://www.ogf.org/gf/event_schedule/?id=2029
On 03/18/2010 05:15 PM, Andrea Cristofori wrote:
Dear all,
I have uploaded the slide I presented in Munich last Monday.
As soon
as
we check the minute of the discussion I will upload the file,
we can
continue the discussion and decide how to proceed.
Cheers Andrea
On 03/10/2010 04:40 PM, Andrea Cristofori wrote:
Dear all,
We have a session scheduled for Monday, March 15 from 4:00 pm
to
5:30 pm
Location: HGB-A 016.
Agenda: - UR for Storage Accounting discussion - AUR discussion - AOB
If you want to add or change something on the agenda please
let me
know.
I'm plannig to prepare some slide to show to the people
present
there
that can be a starting point for the discussion.
Cheers Andrea
-- Andrea Cristofori INFN-CNAF Viale Berti Pichat 6/2 40127 Bologna Italy Tel. : +39-051-6092920 Skype: andrea-cnaf
-- ur-wg mailing list ur-wg@ogf.org http://www.ogf.org/mailman/listinfo/ur-wg
-- Scanned by iCritical. -- ur-wg mailing list ur-wg@ogf.org http://www.ogf.org/mailman/listinfo/ur-wg
-- ur-wg mailing list ur-wg@ogf.org http://www.ogf.org/mailman/listinfo/ur-wg
-- --- Dr. Johannes Reetz ---------------------------------------------------------- Rechenzentrum Garching (RZG) der Max-Planck-Gesellschaft / MPI für Plasmaphysik Garching Computing Centre of the Max Planck Society / MPI for Plasma Physics Boltzmannstrasse 2, D-85748 Garching, Bldg D2 office #315 tel +49-89-3299-2199 fax +49-89-3299-1301 http://www.rzg.mpg.de