- To have a common perspective on the forward look document (including what really needs adding before we can start editing);
- Agree on slides needed for the Tier-2 session (to motivate discussion and make the current and emerging situation clear to anyone not closely involved) and how best to distribute the preparation.
- Capture some key questions we should seek to answer in the discussion session.
Early thoughts on session:
1) We have/are evolving the Tier-2s towards a few big sites with storage and other sites contributing CPU (only.... though that doesn't mean they don't have storage). This is something we have already agreed/implemented and is in-step with WLCG evolution. There was the recent presentation from the WLCG working group about cost-models, which demonstrates (amongst other things) the potential savings be
rationalising storage. Perhaps include a summary of this work.
2) Aside from the cost arguments, there is also a technical discussion about how this is implemented. There is the future "Data Lake" thing, for example, and caches and federated data etc.
3) In GridPP6 we have to be clear about the role of staff we propose to deploy at Tier-2s: They are definitely not "tier-2 hardware post"; they are skilled RSE/RIEs . So we sort of need a job description almost.... and the purpose of all this is to help crystalise
the arguments as to why they provide essential value.