cubicweb #17057223 reimplement looping tasks management [validation pending]
When started as a wsgi application from, e.g., gunicorn, we cannot start looping tasks as usual. So we need an alternative to the current tasks manager that could run as a separate process along with those of the web server.
This could be done by relying on an external scheduler (something like cron, but running along with the cubicweb application). Possible candidates are:
The latter looks simpler but the former has builtin support for multi-tasking execution (defaulting to a thread pool executor model).
|closed by||#f7ff5217a02f [server] replace repository's tasks manager by a scheduler|
|patch||[server] introduce a scheduler class to run repository "looping tasks" [applied][server] Add a "scheduler" command to run repository scheduler [applied][server] replace repository's tasks manager by a scheduler [applied][server] introduce a scheduler class to run repository "looping tasks" [rejected][etwist] Do not call repository's start_looping_tasks anymore and warn about this [applied][pyramid] Drop call to repository's start_looping_tasks method and warn about this [applied][wsgi] Drop calls to repo.start_looping_tasks [applied]|