I totally agree that debuging such things can be difficult. I just wonder if there are some mechanisms that can kill tasks before they are finished. 100h process let be honest - it is not a common case. Honestly speaking i am not pretty sure how frequent this problem can be. I did not manage to perform enough tests :P
My main suspect is the systemd. I try to figure out if it possible to create some kind of listener to get the source of the SIGTERM (strace utility??). For some reason i think graphical.target can be a reason: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=825394 My next step will be to refactor the code to run with only *multi-user.target *dependency. I need somehow to rebuild the gstreamer pipeline where the preview is realized by second, independent application. I will come back when I get any improvement or conclusions. -- Resources: - http://web2py.com - http://web2py.com/book (Documentation) - http://github.com/web2py/web2py (Source code) - https://code.google.com/p/web2py/issues/list (Report Issues) --- You received this message because you are subscribed to the Google Groups "web2py-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
