[BLAST_SHIFTS] What to do if CODA fails to end, but run is good

From: Samuel P. Hariton (shariton@brandeis.edu)
Date: Wed Jul 16 2003 - 17:52:17 EDT


Hey,

Its happened a couple of times that for one reason or another the final
checklist window won't run. (Coda crashing is probably the most common
cause of this).

If this happens, it means that the entry in the RUN table of the
BLAST_physics database will list the data and configuration as
"unconfirmed", and no submission will be made to the webpage. To prevent
this, when there is a problem, run "~elog/bin/check_end.tcl" from
blast@dblast07 _BEFORE_ starting the next run (if you start the next run,
then it will likely cause problems because the script will be confused
which run it is supposed to be on. This will bring up the end of run
window. (If it doesn't, then its probably a bug with check_end.tcl itself,
at which point I will hunt it down and correct any problems).

Not running check_end.tcl won't cause any loss of data, but it will
probably keep any automated crunching programs from knowing whether or not
that run is worthwhile.

 - Sam Hariton



This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:29 EST