2.3.3 Describe Azure Storage Redundancy
Description of Azure storage redundancy types.
Last updated
Description of Azure storage redundancy types.
Last updated
Azure Storage keeps multiple copies of data to protect from planned & unplanned failure. When deciding which redundancy option is best, consider the following factors:
How data is replicated in your primary region
Whether it is replicated to a second, geographically distinct region
Whether your app needs read access to the replicated data in the secondary region if the primary region fails
Data in an Azure Storage account is always replicated three times in the primary region. In the primary region you can choose LRS (Locally Redundant Storage) or ZRS (Zone-Redundant Storage).
LRS replicates data three times in a single region. This provides at least 11 nines of durability (99.99999999999%) over a given year. This is the lowest cost and least durable of the options. If a fire or flooding occurs in the datacenter, LRS replicas may be lost or unrecoverable. To mitigate this, MS recommends using ZRS (zone-redundant storage), geo-redundant storage (GRS) or geo-zone-redundant storage (GZRS).
In Availability Zone-enabled regions, ZRS replicates Azure Storage data synchronously across three Availability zones in the primary region. ZRS offers durability for at least 12 nines (99.999999999999%) over a given year. Even if a zone becomes unavailable, ZRS replicas are still accessible for read/write operations. It is recommended to use ZRS in the primary region for high-availability scenarios. ZRS is also recommended for restricting replication of data within a country/region.
Data can be replicated to the secondary region in your Azure Region Pair, providing durability in the event of a disaster in the primary region. The two options for this are: GRS (Geo-Redundant Storage) and GZRS (Geo-Zone-Redundant Storage). By default, data in the secondary region isn't available for read/write unless there is a failover. Once failover is complete, the secondary region will be available.
As data is replicated asynchronously to the secondary region, an unrecoverable failure of the primary region may result in data loss. The interval between most recent write to primary and last write to secondary is known as the RPO (Recovery Point Objective). Azure Storage typically has an RPO of less than 15 mins.
GRS copies data synchronously three times in a single location in the primary region using LRS. This is then asynchronously copied to a single physical location in your region pair with LRS. GRS offers 16 nines (99.9999999999999999%) of durability over a given year.
GZRS combines high availability provided by redundancy across availability zones with protection from regional outages provided by geo-replication. GZRS data is copied across three Azure availability zones in the primary region (similar to ZRS) and is also replicated to a secondary geographic region for protection from regional disasters. GZRS is recommended for maximum consistence, durability and resilience. GZRS is designed to offer 16 nines (99.9999999999999999%) of durability over a given year.
GRS or GZRS replicate data to another physical location. That data is available to be read-only if a failover is initiated from primary to secondary. If you enable RA-GRS or RA-GZRS then the data is always available to read, even when the primary is running optimally.
