Scaling apps to zero containers on Enclave nows redirects your traffic to Enclave’s error-page server (Brickwall) before shutting down app containers.

Concretely, this means the failover from your app to your Custom Maintenance Page (if you configured one) will happen smoothly: clients will never see a generic error page.

For comparison, if you scaled down to zero containers before this change, the failover would happen automatically, but only once our monitoring detected your app was down. Often, this resulted in about a minute of latency during which clients would indeed see a generic error page.