Speaker
Andrew McNab
(University of Manchester)
Description
Some of the new STFC user communities we are talking to want to use High Throughput Computing resources like ours but with very different interfaces. e.g. they want to use OpenStack and orchestrate VMs themselves which then talk to each other within the site. How should we respond? How would that affect GridPP sites? What is feasible for us to do? Can conventional approaches be sufficient in practice?