On 16 Nov 2007, at 17:19, Burke, S (Stephen) wrote:
> Testbed Support for GridPP member institutes
>> [mailto:[log in to unmask]] On Behalf Of Graeme Stewart said:
>> Because PanDA stages input datasets on the site's SE and
>> puts outputs onto the site's SE as well (it does all other data
>> movements asynchronously using ATLAS DDM) we will not see
>> these large
>> data management timeouts which currently cripple atlas production in
>> EGEE.
>
> What does it do if the put to the close SE fails?
It fails.
Or, at the site's discretion it will leave the outputs on the worker
node (or in a shared "atlas jobs failed" area) and the next pilot
will attempt to recover the outputs. Some OSG sites were nervous
about this, and had it turned off, but have later realised it's good
for them, because they get higher job success rates, and have
subsequently enabled it. It helps buffer you over glitches with local
storage.
> And does it still send
> the log files to CERN?
No. They are stored on the local SE and there's a nice system to
srmcp them to BNL on demand if you want to browse them (SRM over http!).
Cheers
Graeme
--
Dr Graeme Stewart - http://wiki.gridpp.ac.uk/wiki/User:Graeme_stewart
ScotGrid - http://www.scotgrid.ac.uk/ http://scotgrid.blogspot.com/
|