AWS's own cloud-optimized relational database — a supercharged version of MySQL and PostgreSQL built specifically for the cloud.
The fundamental difference is in how data is stored and replicated.
Each replica has its own separate storage. When the master updates, the data must be physically copied to every replica — this takes time and creates lag.
Master → Copy data → Replica 1 (slow)
→ Copy data → Replica 2 (slow)
→ Copy data → Replica 3 (slow)
All instances (master + replicas) point to the same shared storage volume. No copying happens — replicas see changes instantly.
Master writes to Shared Storage
↓
All replicas read from the SAME storage
↓
Everyone sees updates instantly — no lag
YOUR APPLICATION
|
+----------+----------+
| |
WRITER ENDPOINT READER ENDPOINT
(Always → Master) (Load balanced)
| |
v v
+--------+ +----+----+----+----+----+
| Master | | R1 | R2 | R3 | R4 | R5 |
+--------+ +----+----+----+----+----+
| | | | | |
v v v v v v
============================================
| SHARED STORAGE VOLUME |
| Auto-expanding: 10GB → 256TB |
| Striped across 100s of volumes |
| Self-healing + Auto-replication |
============================================
|
+--------------+--------------+
| | |
AZ-1: 2 copies AZ-2: 2 copies AZ-3: 2 copies