Attendance:  

  Stuart, Kent, Peter Shawhan, Linqing, Erik, John, Joe, Jolien, Ben,
  Duncan, Patrick, Alan,  Igor,  Philip.


Action:

  1. Kent wil discuss plans for the next release.  

  2. Philip would like feedback about the resampling in the frameAPI.

  3. Follow up on LDAS at UTB

  4. Jolien will circulate an announcement about LAL code freeze

  5. Next meeting: use of search summary.  

  6. Two weeks from today for run plan padding.


Announcements:

  1. S1:  29 June -- 14 July


Status of LDAS:

  They have all systems up and running with new version of LAL and
  LALWrapper on them.  Everything is up, running and ready to be
  used.  

  Plan another release of LDAS for S1.  It will be 0.3.0.  The plan is
  have that at sites and tested about 1 week before the science run
  starts.  End of third week of June.  Code freeze from June 3.  They
  will have a meeting on Monday at Caltech to decide on priorities.
  Primary task is to modify the way in which API's communicate
  information;  they think this is a big problem with job queuing.  

  RDS.  They are carrying it on their schedule to hash out all the
  problems and resolve the issues.

  Greg is re-generating SFT's again.  At LHO going well.  At LLO he's
  having some problem associated with the timing issues.   File
  transfer get initiated but does not complete before the frameAPI
  needs it.  He will be using them until Monday or so.  First time
  through,  took 3-4 days per interferometer.

  Erik will be using the LDAS system at MIT.  Next week it will be
  used quite extensively at MIT.  Julien is already doing things quite
  extensively.  

  
Status of LDAS at UWM:

  Up since Tuesday noght.  Looking good, and looking stable.  Don't
  push things as hard as the lab system.  He has put up an FAQ that is
  linked from both medusa and computing web pages.  

  Beowulf hardware will be purchased soon.  Stuart will send us a list
  of the hardware so that we can make a similar purchase.  


Status of LAL and LALWrapper:

  CVS is now re-opened.  Greg will be at Milwaukee during the week of
  LDAS camp to get that pulsar code into LAL.  Don't have any real
  major plans for overhaul of LAL and LALWrapper.  

  Are we going to coordinate the LAL and LALWrapper release with the
  next release of LDAS.   Kent would like about three weeks to resolve
  bug issues in the next release. 

  10 June 2002 as tentative code freeze date.  


Status/Plans for DSO's:

  1. Stochastic:  John will take all clean locked stretches to compute
  an expected sensitivity for those stretches.  Plan to run on E7 data
  at UWM.   They will let us know which database.  

  2. FCT: Has not been putting much time for past few weeks.  Will be
  putting more time from now on.  They still need to do some tests.
  Teviet has some time for simulations with signals into the data.  

  3. Inspiral:  Plan to tidy up and use some of LDAS features of latest
  release of LDAS.

  4. Power:  Will run through the E7 data.

  5. SFT:  see above.


Run plans for E8/S1:

  LDAS side technical point.  Planning to change all the passwords
  after the next release.  People should be aware of it.  

  48 nodes at each of LHO and LLO with 512MB Ecc,  2GHz P4 single
  processor fast ethernet, minimal disk. 

  We should come up with the initial plan.  We should pass it to the
  computing committee.  

  Utilization of computing resources continued after the E7.  This is
  an important place for us to state what needs to be done in terms of
  data, etc.  How much time to move around?  

  There are already people in the control room at all times.  So we
  can do this.  Whatever we like that does not interfere with the
  plans of the commissioning people.  The only goal is that we should
  production of reduced data sets.   Minimize burden on operators.

  Which search codes need response?  inspiral, FCT, ringdown.  

  Getting clean segment information.  DMT which checked stored power
  in the arms.  Reliable if it is tuned correctly.  There is no
  problem with that except getting information into the database.
  Afterwards,  the work was to check that the controls were not
  changed.   He does not know if the channel was created to record
  being into detection mode.  What are the chances of 

  Do we have any software or OS upgrade plans before S1?   

  They will switch to RH7.3 (or 8.X depending on what happens) and
  solaris 9.


AOB:

  What is the story of the next release of the frame library?  Benoit
  will be working on it more quickly once things get under control.
  Albert sent around what is the latest snapshot.  Not implemented by
  LDAS.  To properly handle frequency series in frame spec,  they need
  to tweek a few things.   People should contact Albert for the latest
  copy version that is available.   Plans from CDS will not upgrade to
  the newest specification.  They promised that it will be done by
  S2.  The frames that will be built from V4 frames from CDS.

  The arm azimuth is wrong in the frames for E7.   Peter will
  circulate an e-mail to the UL Chairs to make sure they are aware
  that some of the data is wrong in the frames.

  Any changes needed.  Albert is taking the lead. 

  Eventually,  Joe wants to run Monte-Carlo simulations.  They want to
  add simulated data to data in datacondAPI.  The way datacond is
  right now,  you cannot add two timeseries and carry appropriate
  meta-data.   Early june is ok with that.


Next Meeting:

  Thursday, 9 May 2002 at 10:30 PDT