[Condor-devel] starter wisdom: what happens when a job exits?

[Derek Wright <wright@xxxxxxxxxxx>]

As per previous threads on this list, I've spent a lot of time in the  
past few months adding/changing a ton of code in the startd and  
starter so that there's now a system of hooks in place at various  
stages of a job's life-cycle on the execution machine.  One of the  
things I had to do was come to grips with what the starter actually  
does when a job exits (which was much more challenging than it  
sounds). ;)

As a public service to the Condor development community (and to help  
myself remain sane while I had to change some of it), I wrote up a  
fairly detailed but high-level explanation of what's going on in the  
starter code whenever a job exits.  The results live in the "WISDOM"  
file in the src/condor_starter.V6.1 directory (in HEAD).  The write- 
up is an explanation of how things are in 7.1.* and beyond with the  
hooks in place -- I didn't bother to write it all up as the old work- 
flow with the older (if can believe it, more insane) names, etc.

If you're interested in this topic and have any questions, please ask  
me sooner rather than later so I can edit the document while a) I  
still have time, and b) it's still mostly swapped into my working set.

Cheers,
-Derek