Sysadmin Central

A shorter post for this week, let’s cover something that I think most of us have a decent understanding of but that’s always worth being confidently knowledgeable about, data redundancy.

Azure offers a range of redundancy options to help you safeguard your data against failures but it’s very easy to overlook such an important setting whilst configuring new services or whilst reviewing existing services.

In this post, I’ll briefly break down each of the available redundancy strategies available in Azure, highlighting when and why you may want to use each level.

Locally Redundant Storage (LRS)

LRS keeps three copies of your data within a single physical location in the selected Azure region. While LRS provides a good level of protection against individual hardware failures such as harddrive faults, it doesn’t protect against a data center outage and could result in a total loss of data in the event of a localised event.

LRS is a cost-effective solution when your data doesn’t need to be replicated across regions. It’s often sufficient for scenarios where data loss is acceptable, such as in development or testing environments.

Zone-Redundant Storage (ZRS)

ZRS stores data across 3 availability zones within the same region, in entirely seperate data centers, ensuring that a data center failure in one zone doesn’t result in data loss.

ZRS is ideal for production workloads that require high availability and must withstand data center failures, but where replicating data to another geographic region isn’t necessary. Use it when local fault tolerance is a priority but cross-region redundancy is not a must.

This level tends to be the best balance of cost vs data protection unless you have specific requirements than require a greater level of redundancy.

Geo-Redundant Storage (GRS)

GRS replicates your data to another region hundreds of miles away from the primary location. With GRS, Azure stores three copies of data in the primary region (LRS) and then asynchronously replicates another three copies to a secondary region.

GRS is critical for disaster recovery scenarios where your data must be preserved even if an entire region becomes unavailable. It’s particularly valuable for business-critical applications and compliance scenarios requiring off-site backups but is more expensive than the lower level options.

Read-Access Geo-Redundant Storage (RA-GRS)

RA-GRS extends GRS by allowing read access to the secondary region’s data. While GRS provides data availability in a secondary region only after a failover, RA-GRS allows you to read from that region even before a failure occurs, which is great when you need higher data availability across regions.

This redundancy level is especially useful for global applications that need to provide low-latency read access to users in different geographical areas.

Choosing the Right Redundancy Level

The choice of redundancy level depends on your business needs, cost considerations, and tolerance for downtime or data loss. LRS is suitable for less critical workloads, while ZRS, GRS, and RA-GRS offer increasing levels of resilience and are appropriate for higher availability and disaster recovery requirements but do come at a potentially significant cost.

When architecting your Azure solution, ask yourself:

• How critical is this data?
• Can the data be easily recreated if lost?
• Can the data afford to be unavailable for a short period (minutes or hours)?

Understanding these redundancy levels allows you to build a robust infrastructure, aligned with both your budget and availability needs, ensuring your services remain active even in the face of unexpected failures.

Hopefully this short post clarifies when and why you may use the redundancy levels, if you’re still unclear or want a little bit more detail then the best place to look is always the official Microsoft documentation – Data redundancy – Azure Storage | Microsoft Learn