Paul, As you know, XSEDE has a partial implementation of GLUE2-JSON in production. Our source code is available in GIT at: https://github.com/ericblau/ipf-xsede/ <https://github.com/ericblau/ipf-xsede/> Step 1, IMO, is for ATLAS to look at the GLUE2 data model and figure out how the information in their JSON file maps on to the GLUE2 reference model. Many GLUE2-WG members can help with that if they can’t do it on their own. Step 2 would then be to look at existing implementations, such as XSEDE’s, and decide whether to extend that implementation with the information they want, or to develop something new. XSEDE is available to discuss what it would take to add specific types of information if they can identify what those are. Perhaps they could share documentation and examples on those files for others to help with the above steps? JP
On Oct 14, 2016, at 1:23 PM, Paul Millar <paul.millar@desy.de> wrote:
Hi all,
Thanks for the feedback so far.
Just to clarify what's happening here: ATLAS has decided to do storage accounting by requiring sites/storage to have a JSON file written a well-known location. This is not under discussion.
They have formulated a custom/proprietary JSON format that includes all the information they want to know.
The proposed mechanism and JSON file format has been presented in the September pre-GDB:
https://indico.cern.ch/event/394833/#3-resource-reporting-proposal
I suggested they consider GLUE2-JSON instead. To do consider GLUE2-JSON, they need:
1. some people with which they can discuss any issues,
2. example JSON files that provide complete examples.
I suggest we tackle 1. first and look at 2. as an outcome of the people in 1.
So, simple question to everyone in this group:
Who wishes to help ATLAS evaluate GLUE2-JSON?
Cheers,
Paul. _______________________________________________ glue-wg mailing list glue-wg@ogf.org https://www.ogf.org/mailman/listinfo/glue-wg