- Ably FAQs
- Performance and Redundancy
- Redundancy & reliability
-
Account billing and packages
-
General
-
Realtime API and client libraries
-
Troubleshooting
-
Channels
-
REST API and client libraries
-
Ably error codes
-
Account and app setup and configuration
-
Ably architecture, transports, and security
-
Performance and Redundancy
-
Push Notifications
-
Integrations
-
Migrating to Ably from an existing service
Self-healing cluster
The Ably realtime system uses a consensus algorithm to communicate between all servers and agree which services are healthy and which are failing.
Any issues in the cluster will therefore:
- Be isolated from the healthy parts of cluster as the majority agree on the state of the unhealthy servers
- Rectified automatically by our monitoring system and health servers that intelligently fix and replace servers as required