> I see the documentation for issuing a SIGHUP for restarting gracefully. > Is there a similar version of shutdown? I would use this in my > graceful/safe patching system to fire up the zergs, gracefully restart the > main process then gracefully shutdown the zergs. We have some backend > transactions, like large file upload from a cellphone, that can take > hours, > so have to have a nice process for this. Right now I am doing a graceful > stop and parsing the result to know when all have been suspended, then > issue a kill. Is the suspend (SIGTSTP) graceful as well? I.e. if we > issue that in the middle of a 2 hour upload, will everything continue to > process until the mercy timeout or will it just be a hard cutoff?
SIGINT/SIGQUIT are graceful > > BTW: from experimentation the touch-chain-reload option does not appear to > actually load the new code. > it requires application loaded after fork() via --lazy-apps, otherwise the code from master will be copied -- Roberto De Ioris http://unbit.it _______________________________________________ uWSGI mailing list [email protected] http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
