Notejoy's production database is setup with full master/slave replication, enabling hot failover immediately if any issues occur with Notejoy's primary database. In addition to that, daily geographically distributed offline backups are taken of our production data to provide even further redundancy.
Notejoy runs many app servers at a time, with traffic load balanced between them. If any app server is experiencing issues, it is automatically removed from the rotation and traffic is sent to the remaining healthy servers. This provides significant redundancy for Notejoy's app servers as well.
Daily deployments are done on a rolling basis across app servers to also ensure they do not cause any downtown for the service.
Notejoy also has 24/7 monitoring and alerting to ensure the team immediately addresses any availability or performance issues with the service.