Skip to content

Latest commit

 

History

History
52 lines (41 loc) · 1.2 KB

RDS.md

File metadata and controls

52 lines (41 loc) · 1.2 KB

Multi-AZ failover

  • MySQL, Oracle, PostgreSQL -> replication
  • SQL Server -> mirroring

Failover -> RDS DNS endpoint doesn't change

what for:

  • HA
  • Backups from secondary
  • Restores are take from secondary
  • Not a scaling solution
  • For scaling use read replicas

You can force failvoer by rebooting instance. RebootDBInstance API call

Read Replicas

scale out for read-heavy workloads, async, CreateDBInstanceReadReplica API

  • scaling beyond compute or I/O capacity of a single DB instance
  • reporting/data warehousing on replicas
  • have own DNS endpoint
  • could be promoted to standalone DB
  • up to 5 read replicas for MySQL, PostgreSQL & MariaDB
  • could be in different regions
  • async only
  • supported for multi-az
  • RR can't be multi-az
  • chained replication is allowed
  • snapshots/automated backups from RR
  • ReplicaLag cloudwatch metric

engines

  • MySQL
    • native sync
  • PostgreSQL
    • native sync
  • MariaDB
    • native sync
  • Aurora
    • SSD backed virtualized storage layer, replicas share the same storage

Creating read replicas

if !Mutli-Az
    snapshot of primary DB, brief I/O suspension for ~1 minute
else
    snapshot of secondary DB, no performance hits for primary