Write cron business logic as a Python class and let this app do the rest! It enables Django projects to schedule cron tasks, tracks their success / failures, manages contention (via a cache) etc. Basically takes care of all the boring work for you :-)
Currently the problem is that failure logs only go to the database, but there is no possibility to route failures also to the loggers. Having failure logs go to the logger system allows routing them to things like Sentry for easy monitoring.
Coverage increased (+0.02%) to 89.78% when pulling 45fd7404f7d233196f9d3f2e39f994f89f5c7174 on jaywink:log-failures into 225f36204be2d01ef9fd6f80345e7a4812834842 on Tivix:master.
Currently the problem is that failure logs only go to the database, but there is no possibility to route failures also to the loggers. Having failure logs go to the logger system allows routing them to things like Sentry for easy monitoring.