The current implementation requires a system-level cron job to restart the containers each day. This is hacky and fragile, and the most recent redeployment changed the container names, so our cron job restarts weren't working. We should simply make the ZTF consumer roll over to the next day's stream on its own.
The current implementation requires a system-level cron job to restart the containers each day. This is hacky and fragile, and the most recent redeployment changed the container names, so our cron job restarts weren't working. We should simply make the ZTF consumer roll over to the next day's stream on its own.