This is a partial fix to filling the rpi SD card with data. It enables automatic couchdb database and view compaction on a schedule.
The PFC 2.1 running a recipe writes about 1.5G per day to couchdb. This change compacts the data and makes the daily hit about 100M.
Not the real solution, but just delaying "death by filling the disk" until the end of a recipe.
The real solution involves a back end where data is replicated/archived to, and a way to (cron script or from a ROS node that starts a new recipe?) that trims the environmental_data_point DB.
This is a partial fix to filling the rpi SD card with data. It enables automatic couchdb database and view compaction on a schedule.
The PFC 2.1 running a recipe writes about 1.5G per day to couchdb. This change compacts the data and makes the daily hit about 100M.
Not the real solution, but just delaying "death by filling the disk" until the end of a recipe.
The real solution involves a back end where data is replicated/archived to, and a way to (cron script or from a ROS node that starts a new recipe?) that trims the environmental_data_point DB.