Redundancy Options¶
Azure Storage provides multiple redundancy options to protect data from planned and unplanned events.
Note
Secondary-region replication is asynchronous for GRS/GZRS and can have recovery point lag.
Comparison Table¶
| Option | Copies | Regions | Availability Zone | Failover | Durability | Availability SLA |
|---|---|---|---|---|---|---|
| LRS | 3 | Single | No | No | 11 nines | 99.9% (Cool 99%) |
| ZRS | 3 | Single | Yes | No | 12 nines | 99.9% (Cool 99%) |
| GRS | 6 | Two | No | Yes (Customer-managed) | 16 nines | 99.9% (Cool 99%) |
| GZRS | 6 | Two | Yes | Yes (Customer-managed) | 16 nines | 99.9% (Cool 99%) |
| RA-GRS | 6 | Two | No | Read from secondary | 16 nines | 99.99% read / 99.9% write |
| RA-GZRS | 6 | Two | Yes | Read from secondary | 16 nines | 99.99% read / 99.9% write |
Redundancy Topology¶
```mermaid graph TD Data[Data Ingest] --> Primary{Primary Region} Primary --> LRS[LRS: 3 copies, 1 DC] Primary --> ZRS[ZRS: 3 copies, 3 AZs]
LRS --> GRS[GRS: Async copy to Secondary]
ZRS --> GZRS[GZRS: Async copy to Secondary]
GRS --> Sec[Secondary Region]
GZRS --> Sec
Sec --> RA[RA: Read Access to Secondary]
```