Good idea to look through uwsgi.log. The problem was harakiri on a process after 60 seconds. There was an entry in the log along the lines of "HARAKIRI ON WORKER 4 (pid: 2557, try: 1)". To fix, I increased harakiri to 120 in /etc/uwsgi/web2py.xml.
That created a new problem which showed up in uwsgi.log as "SIGPIPE: writing to a closed pipe/socket/fd (probably the client disconnected) on request ". Turns out this is because by default nginx waits a max of 60 seconds for uwsgi to respond to a request. The fix was to add a line to /etc/nginx/sites-available/web2py saying "uwsgi_read_timeout 120;" Helpful reading for anyone else who comes across this same problem: http://stackoverflow.com/questions/24133315/fixing-broken-pipe-error-in-uwsgi-with-python Thanks again for having pointed me in the right direction. Toby -- 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.
