[Henrik Bengtsson] >I am looking for a way to have one R session lock a file for >read/write access, while being updated/modified by another R session. >This will provide me with a-poor-mans parallelization method. It is >ok to have so called advisory looking (as in Unix), which are >non-mandatory to follow. If not available, I'll use lock files, but >there are some potential problems in creating such in an atomic way. >Ideally I wish to have this working on all platforms.
Something which worked for me, so far, is to create a directory, and see if the creation succeeded or failed. If the creation succeeded, you have the lock. If it failed, something else created it and has the lock. You then keep retrying after some delay. Of course, the directory should not be used outside your locking protocol. On the few platforms I used, but likely on most of them, creating a directory "atomically" ought to be safe. Yet, this might be a bit heavy-weighted when locks are meant to be light and quick. If you want some more sophistication, leave a file within the lock directory (once you successfully created it and so, acquired the lock), telling who you are. Destroy the file before destroying the directory (of course). More typically on Unix, put in this file the id of the running process, so when creating the directory fails, the attempter may look in the file, and if there is no such process running, consider the lock stale, and preempt it (to do this safely requires some thinking). -- François Pinard http://pinard.progiciels-bpi.ca ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
