Attendance:  

  Joe,  Warren,  John,  Duncan,  Stuart,  Teviet,  Ed,  Jolien,  Greg,
  Isabel,  Erik.  


Preparations to run DSO's during S1:

  Plan is to run all scripts from a single web-accessible location at
  each of LHO and LLO.   We have to leave this until the last minute
  since things are not ready yet.   The designated machines will be
  sand and delaronde which are used for DMT: John z.  and Szabi are
  the contacts.

  ONLY 10 jobs can run at a time under LDAS.  This is the maximum
  number of assistant managers.   There is high overhead to do logging
  information.   If we really want to,  we can get some more.   Kent
  believes things will work up to 20 assistant managers.

  Will need to e-mail people about using this web page,  etc.
  Situations when everything is going wrong ..... please call greg.  

  Discussion of possible node requirements:  
    IUL - AS_Q,  4  up to 5 other
    BUL - Slope:       4 nodes           Power:       4 nodes |  ->  unaware of any more = 
          TFClusters:  4 nodes /

  Allocate some nodes and investigations.  A question arose whether the deafult
  database at the sites should be the science database.  There was
  general agreement that it should not,  but there was concern that
  other software may assume it is.   Therefore,  no change to past
  configurations will be made.  


  Erik:  testing of use of RDS's ....... one or two nodes to run
  something on RDS's   should we do this ...... nobody has used RDS's.
  We really do need to test this.  There is no point to doing it if
  there is   Stick in as slot on our plan.   

  Stuart mentioned a consideration for S2,  etc:  do we always want to
  run with such small numbers of nodes.