Workload architecture
AOSREL02: How do you provide high availability for your OpenSearch Service domain? |
---|
To protect your OpenSearch Service domain during a service disruption, you can deploy nodes across two or three Availability Zones within the same Region. This configuration, known as Multi-AZ, leverages isolated locations called Availability Zones within each AWS Region. You can also choose to deploy OpenSearch with Multi-AZ with Standby deployment to protect against infrastructure failures. Enabling Multi-AZ with Standby deployment increases the availability of your domain to 99.99%.
AOSREL03: How do you monitor quotas and limits in your OpenSearch Service domain? |
---|
Cloud providers have built-in usage constraints that prevent excessive resource utilization. Understanding and monitoring these quotas is essential to prevent scaling issues and ensure uninterrupted service operations.
AOSREL04: How do you recover your OpenSearch Service domains from outages? |
---|
Disaster recovery involves plans and processes to resume operations and restore data in the aftermath of a disruptive event or catastrophe. Having a disaster recovery plan is important to recover from any disruptions to the OpenSearch domain.
AOSREL05: How do you use reliable instances for your production search workloads? |
---|
Business-critical workloads demand enough resources in production environments to handle reliability, stability, and high performance. In contrast, dev environments can function on minimal resources without impacting critical operations.
Best practices
AOSREL02-BP01 Create Multi-AZ with Standby OpenSearch Service domains
AOSREL03-BP01 Regularly review your OpenSearch Service quotas
AOSREL04-BP03 Enable index replication for your critical indices
AOSREL04-BP04 Employ cross-cluster replication to achieve higher availability
AOSREL05-BP01 Implement appropriate compute sizing for production workloads