Choose timezone
Your profile timezone:
GridPP Technical Meeting
→
Europe/London
Virtual Only
Virtual Only
,
Description
Fortnightly meeting for technical topics looking further ahead than the weekly ops meetings on Tuesdays. There are also dedicated storage group meeting on Wednesdays. Each topic can go beyond the nominal 5 minute slot whenever necessary.
Brief Notes
The purpose of the meeting was to stake stock and decide how to run these meetings. The agenda was therefore largely a red herring.
Storage
Largely due to lack of experiment engagement we are behind in the testing. So Sam is thinking about doing things at a GridPP level. So for example sharing storage (via xrootd federation) between reliable and less reliable site, e.g. UCL has storage at QMUL and perhaps testing within ScotGrid might be an option. Other testing with Atlas is also planned.
There are at least two xrootd implementation on ceph. Better integration with Rucio would be good.
Compute
The Liverpool experience with vac is very encouraging. Andrew is pushing to increase the number of VMs available e.g. CMS running again. All the GridPP VOs work? ALICE is also comingon board. HTCondor VMs are also being worked on which will benefit many of the VOs (Atlas, CMS and ALICE at least).
People are talking about federated OpenStack instances. This is coming from the Scientific Working Group.
SKA meeting next week will have a significant OpenStack component and we could learn from this. [ Do we need to pay for OpenStack installations rather than DIY? Does this (installation) count as capital? ] We need to point out that OpenStack is a platform and we need to build on top of it. We should suggest GridPP Dirac as part of this.
Services
Dirac services at Imperial. Possibly monitoring and analytics. Currently provided by CERN but maybe we could provide for others... anybody interested? Accounting is also part of this. Moving of Data? Perhaps interfacing with FTS (being worked on at Imperial for NA62 and MICE but generic).
Network
Generally GridPP sites work OK.
Security
No change of direction. Most work is around the WLCG working groups. David Crooks is the chair of one these groups and this work is ongoing. There was also work on traceability. Operationally this needs.
Others....
Some experiments are looking at bursting data for several hours .. a case for a Science DMZ?
There are minutes attached to this event.
Show them.
