DataBases High Availability

Choose the right database

High Operational Efficiency and Less Control

• DynamoDB (no sql)
• Aurora (sql)

Medium Operational Efficiency and Control

• RDS (sql)

Low Operational Efficiency and Control

• EC2 (sql)

High Available DynamoDB

• Distributed data and incoming traffic across partitions by default
• Partitions are replicated across multiple AZs
• Global tables allow multi-region availability

Multi-AZ RDS

• Synchronous replication across multiple AZs
• Failover to standby in case of primary failure
• Automated backups and snapshots

Regional Read Replica

• Asynchronous replication across regions
• Read-only copy of the primary database
• Can be promoted to primary in case of primary failure
• Warm standby solution

Snapshot Recovery

• Multi AZ RDS cluster
• Automated backups and snapshots

Aurora

• Multi-AZ deployment
• Data copies across multiple AZs
• Automated failover

Aurora Global Database

• One primary region and up to five read-only secondary regions
• Data replicates from primary to secundary regions
• Leverages storage-level replication
• Promote a secondary region to primary in case of primary region failure

Redshift High Availability

• Only RA3 instances supports Multi-AZ
• Best option for HA otherwise is a multi-node cluster
• single-node failure must be restored from a s3 snapshot

Output

• DynamoDB is the best option for high availability
• Aurora is the best option for high availability with SQL
• Redshift is the best option for data warehousing high availability