develop autmated dropper between RefDB and PhEDEx

[ericvaandering]

develop autmated dropper between RefDB and PhEDEx

[ericvaandering]

Comment by lat on Thu Jul 29 11:46:23 2004

Mostly done now -- RefDBReady. Takes list of assignments, or patterns of dataset.owner (in file or command line). If the patterns, figures out the assignment numbers. Then for each assignment generates the drops. Then checks if files are already ready in castor, and marks assginment directories ready if they are (plus touches go).

Remaining to do: keep track of what's already been sent. Right now all drops are always recreated, and we don't have easy way of comparing what was "ready" the previous time, or was sent with an agent. Also should probably make the "check if all files are on castor" somehow a separate tool, or option, or something, so drops don't need to re-generated.

[ericvaandering]

Closed by barrass on Tue Aug 3 07:21:41 2004