Azure provides four main data redundancy options for its storage accounts, which are defined by how many copies of your data are created and where those copies are stored. The key difference is the level of protection each option offers against different types of failures, from a single disk failure to a regional disaster.
1. Locally Redundant Storage (LRS)
- Replication: LRS is the most basic and least expensive option. It replicates your data three times within a single data center in the primary region.
- Protection: It protects against drive and server rack failures, but offers no protection against a data center-wide outage or a regional disaster.
- Best Use Case: Non-critical data, development and test environments, or temporary data that can be easily recreated.
2. Zone-Redundant Storage (ZRS)
- Replication: ZRS replicates your data three times synchronously across three separate Azure Availability Zones within the primary region. Each Availability Zone is a distinct physical location with independent power, cooling, and networking.
- Protection: This provides excellent protection against data center outages and zonal failures. If one entire data center or zone goes down, your data remains available in the other two.
- Best Use Case: Applications that require high availability and are resilient to a zonal failure, but do not require protection against a regional disaster.
3. Geo-Redundant Storage (GRS)
- Replication: GRS first replicates your data three times using LRS in a single data center in the primary region. It then asynchronously replicates an additional three copies to a single data center in a secondary region, which is hundreds of miles away.
- Protection: This provides robust protection against a full regional outage. If the primary region fails, your data is still secure and can be recovered from the secondary region.
- Best Use Case: Mission-critical data that requires a high level of durability and protection against regional disasters.
4. Geo-Zone-Redundant Storage (GZRS)
- Replication: GZRS offers the highest level of durability. It combines ZRS in the primary region with asynchronous replication to a secondary region. This means data is replicated three times across three Availability Zones in the primary region and then replicated three times to a single data center in a separate, distant region.
- Protection: GZRS protects your data against both zonal failures within the primary region and a complete regional disaster, offering the maximum level of data resilience.
- Best Use Case: The most demanding, mission-critical applications that require both high availability and a complete disaster recovery plan.
Detailed comparison of LRS, ZRS, GRS, and GZRS in tabular form
| Feature | LRS (Locally Redundant Storage) | ZRS (Zone-Redundant Storage) | GRS (Geo-Redundant Storage) | GZRS (Geo-Zone-Redundant Storage) |
|---|---|---|---|---|
| ✅ Purpose | Provides redundancy within a single datacenter | Provides redundancy across multiple availability zones in a region | Provides redundancy to a secondary region, for disaster recovery | Combines zone redundancy with geo-redundancy for maximum resilience |
| ✅ Replication Type | 3 copies of data in the same data center | 3 copies across multiple availability zones | 3 copies in primary region + async replication to secondary region | 3 copies across zones in primary region + async replication to secondary region |
| ✅ Durability | High durability within a datacenter | High durability even in case of datacenter failure | Protects against regional disasters | Protects against zone and regional disasters |
| ✅ Use Case | Low-cost option for non-critical workloads | Better for applications needing high availability within a region | Best for disaster recovery across geographic regions | Best for mission-critical workloads needing max durability and availability |
| ✅ Availability SLA | ~99.9% | ~99.99% | ~99.99% | ~99.99% |
| ✅ Cost | Lowest | Medium | Higher than ZRS | Highest |
| ✅ Example Scenario | Development/test environment | Critical apps that must be available even during datacenter failure | Applications requiring geo-disaster recovery | Financial apps or healthcare where max availability and durability is required |
✅ Summary of Use Cases
| Redundancy Type | Best For |
|---|---|
| LRS | Inexpensive data storage with basic durability guarantees. Suitable for dev/test workloads. |
| ZRS | Applications that need high availability inside a region but don’t require geo-replication. |
| GRS | Applications that need protection against regional disasters but don’t require zone-level fault tolerance. |
| GZRS | Critical applications requiring the highest level of availability, durability, and disaster recovery. |
Key Takeaway:
Choose LRS for low-cost, low-risk workloads.
Choose ZRS when zone-level availability matters.
Choose GRS when regional failover is required.
Choose GZRS for mission-critical workloads requiring both zone and geo-redundancy.
Discover more from Technology with Vivek Johari
Subscribe to get the latest posts sent to your email.
