Reliability
The reliability pillar focuses on verifying that your Amazon OpenSearch Service workload performs correctly and consistently when needed. This includes operating and testing the workload throughout its entire lifecycle. This paper provides best practices for building and maintaining reliable OpenSearch Service workloads.
Design principles
-
Implement monitoring and notifications: Set up monitoring and notifications for OpenSearch domain failures to ensure timely detection and response to potential issues.
-
Provide continuous availability: Implement multi-AZ with standby deployment, Index State Management (ISM) policy, index replication, cross-cluster replication, and regular quota monitoring to ensure OpenSearch Service domain availability.
-
Implement a disaster recovery strategy: Develop a Disaster Recovery (DR) strategy for OpenSearch Service, including regular updates to the latest version, and monitoring of software updates.
-
Stay informed about updates: Stay informed about OpenSearch Service update notifications, keep domains updated to the latest version, and monitor software updates for reliability and security.
