Minutes SLAC-ALTAS TDAQ Meeting 26-Oct-2006
-------------------------------------------

Present: Steffen, Su Dong, Sarah (phone), Amedeo, Matthias, Rainer


Sarah:

MST in Manchester officially ended. Serge tried to set up a partition
with dbproxy in TDAQ 1.6.1.  Was able to bring everything up, but
failed to configure.  Tried without dbproxy, also failed.

Manchester farm still available for tests.  But LST will officially
start in November.  A few nodes available.  Testing will start next
week.

Question about the HLT objects that need to be read from POOL files.
It is planned to distribute these files on the local nodes for tests,
but that has not been done for the MST.  Not clear how this is
expected to work without that.

LST will use month of Nov.  Plan to go from 30 to 100-500-1000 nodes.

There is supposed to be a contact person for each slice that verify
that output data are correct.

CORAL: reported on work to make CORAL thread-safe. Currently it is
not.  Prerequisite to fix the atomicity of 'use' and queries.


Steffen:

Farm up and running.  Worried what the lifetime of the farm is in its
new environment.  Mainly the ambient temperature that makes a
difference.  So not helpful to power up only some of the machines.

Question about how to get the list of nodes.  It's not contiguous.
Matthias suggests to create a flat file in addition to the web
interface.

Consider reserving the first two farm nodes for interactive use.  The
SQL server is also on the first nodes.

Set up a CERN hypernews for the farm use.  Make a list of tests.