On 12 Feb 2014, at 14:12, Ewan MacMahon <[log in to unmask]> wrote:
>> -----Original Message-----
>> From: Testbed Support for GridPP member institutes [mailto:TB-
>> [log in to unmask]] On Behalf Of David Colling
>>
>> We are due another technical meeting this Friday. I said that I would
>> advertise it a few days in advance so that people could add agenda items.
>> I know that Duncan would like to talk about the IPV6 and it is a while
>> since we have had a storage summary. Does anybody else have other items to
>> add?
>>
> At the risk of dropping Andrew even more in it, I'd quite like to
> think about Vac again; as I understand some of the cloud work, we
> have ATLAS images around which are capable of logging themselves
> into panda and pulling in work. I'm interested in exploring whether
> we could get them or something much like them running in vac.
This would be a good time to start looking at other VOs, yes.
I’m hoping I’ll have time to go through some of the recent Vac updates,
including the accounting which is now going all the way through the
chain to the central APEL database and EGI portal; and the LHCb
VM contextualisation which is now finally tagged and includes proper
separation of JobAgent and payload using different unix accounts
within the VM.
I’d also like to look at getting it working with the GridPP DIRAC,
and I’ve got the go ahead from LHCb DIRAC to reuse LHCb's Vac
and Clouds VM contextualisation as a starting point for that in
GridPP. If we can get that working, then users of those VOs can
submit to a central task queue which then gives transparent
access to batch, Vac, and Cloud resources.
Cheers
Andrew
--
Dr Andrew McNab, High Energy Physics,
University of Manchester, UK, M13 9PL.
Skype and Google Talk: andrew.mcnab.uk
Web: www.hep.manchester.ac.uk/u/mcnab
|