Problems with the queueing system?

Dear Dr. Bonvin and fellow users,
We reached a new problem in our big system, but now related with the job handling itself. At the beginning of it1 haddock gets stalled trying to build the first structure of this iteration. I noticed that the last it0 csh “job” script in our case includes the first structure of it1, and that haddock tries to submit this script at the beginning of it1, failing while trying to append the “STARTED” flag to an already gzipped NAME_it0_refine_200.out file (we are building 200 it0 structures and using the two default queues in run.cns). This results in being able to get all structures from it1 except from the first one, stalling the whole job. I was able to resume the job by forcing haddock to rebuild it1, but I was wondering if there’s a proper way for making this. Any suggestions?
Best regards,

Hmmmm … strange

I don’t know what your local settings are, but make sure to kill any python process before restarting haddock