Slightly incomplete notes because the actual notes are in the office on a piece of paper.


* DPM whitepaper.  Severl comments; one is about the effort.  Since the effort required to run DPM is less
than other storage solutions - one of the original selling points of DPM - this is an advantage but there is
also effort needed to build support, eg community, site, tier2, etc.  Sort of goes against grid-specific
or at least it needs critical mass.

On the subject of data centre automation, Jens mentioned work by collaboators in the US who were working
on green computing data centres which were local to the green energy and not to where offices were.  Jobs
could suspend although MPI jobs were harder.  Anyway the point was to make use of Redfish/Swordfish/Kingfish
and also TTU linked in information systems thus coming up with a GLUEfish.  As more stuff in the data centre
becomes software defined it should become easier to add layers of resilience (plan for failuure etc.) but
it is still necessary to know about the hardware.  Workshops had been held at supercomputing.

See nsfcac.org for the data centre automation and control.  The company working on the green computing
is called Lancium.

Raul also sent links to the list highlighting the worker's perspective of the remote working.