WLCG Information System Evolution Task Force

Europe/Zurich
513/R-068 (CERN)

513/R-068

CERN

19
Show room on map
Description
Meeting to discuss the evolution of the WLCG Information System

Attended:

Laurence, Maarten, Oxana, Stephan, Kyle, Salvatore, Andrew, Alessandro Paolini, Kate, Julia

 

The meeting outcome:

The discussion was initiated by the fact that sites running ARC had to apply custom patches in order to provide dynamic info in BDII. Dynamic information, like number of pending and running jobs is currently used by ALICE and LHCb. They both confirmed that they are fine with GLUE2. 

The decision was to request the central fix of the ARC problem by ARC development team. Apparently, if the fix is be limited to GLUE2 only, it should be quite straight forward. The ticket for this problem has been submitted a while ago. Alessandra will update it and will follow up on the resolution. Before the fix is provided, sites should not develop and apply any custom patches.

As a common strategy, both LHCb and ALICE will use dynamic info from BDII, unless the site does not run BDII. In  case when BDII is lacking, both LHCb and ALICE confirmed that they have a way around. ALICE currently uses mostly CREAM CEs. In case when dynamic info is not provided through BDII, HTCondor can be used to drive other CEs and to handle dynamic information. Similar approach is already in production at CERN. For LHCb, in order to get rid of the dependency of the dynamic info in BDII, some changes in pilot factory have to be applied. Andrew does not see a problem with it.

Also discussed the situation if the site does not run BDII but still publishes accounting data via APEL. APEL client has to be modified in this case to get time to work conversion factor from the source different from BDII. Andrew suggested to use local site configuration file to store this conversion factor. Another alternative would be GocDB. Maarten suggested to have a flexibility in the APEL client, so that it could iterate through various methods in order to get required information. Julia will follow up with APEL developers.

All required static and semi-static information will be available in CRIC.

Agreed that there is no need for sites to expose info via JSON file.

 

There are minutes attached to this event. Show them.
    • 15:00 15:30
      Overview static/semi-static and dynamic info required for compute resource description 30m
      Speaker: Julia Andreeva (CERN)
    • 15:30 16:00
      Discussion 30m