Microsoft this week announced public preview support for geo-redundancy on Azure Files standard file shares with large file shares enabled. Let’s look at what this is and how it could affect an AVD implementation with FSLogix. But first, some background information may be helpful.
Azure Files IOPS
To start, let’s talk about inputs outputs per second or IOPS. IOPS measures storage performance and represents the number of read and write operations available on a storage account. IOPS are important with FSLogix profiles. If the number of requested IOPS exceeds what’s available on the storage account, requests are throttled, and user experience will suffer.
An Azure Files standard account supports up to 1000 IOPS by default with options for geo-redundant storage (GRS), geo-zone redundant storage (GZRS), locally redundant storage (LRS), and zone-redundant storage (ZRS). Geo-redundant features of a storage account provide the option for replicating data to a second region, a good option for disaster recovery. However, 1000 IOPS is not sufficient to support FSLogix profiles.
Enabling large file shares can increase IOPS on an Azure Files standard storage account from 1000 to 20,000 IOPS, a significant performance enhancement. Enabling large file shares also increases the total throughput and capacity of the file share. The downside to enabling large file shares is that it’s only available with LRS and ZRS storage accounts. Geo-replicated storage accounts GRS and GZRS have not previously supported enabling large file shares.
That brings us back to the public preview feature that allows us to enable geo-redundancy for Azure Files standard storage accounts with large file shares enabled. With this preview feature, it is possible to get the advantages of large file shares, including 20,000 IOPS, 300 MiB/s of throughput, and up to 100 TiB of data on an Azure Files standard share as well as the ability to replicate data to a secondary region.
Highlights of New Azure Files Geo-redundancy Public Preview
Let’s pull out a few highlights that go beyond the IOPS relevance and benefits from this week’s announcement.
Replication is asynchronous and snapshot-based.
With the preview feature enabled, a snapshot is taken every 15 minutes in the designated primary region and replicated to a secondary region. The most recent snapshot available in the secondary region is used in the event of a failover. The recovery point objective (RPO) for Azure Files standard with geo-replication is 15 minutes or longer, depending on the snapshot size and the geo-replication lag between the primary and secondary regions.
Azure Files with geo-replication support customer-initiated failover.
This feature allows the customer to initiate a failover from the primary to the secondary location. For example, users can initiate a failover due to an availability issue in Azure, a configuration issue, or to test disaster recovery processes. Customers are not dependent on Microsoft to initiate a failover.
Geo-replication for Azure Files standard file shares with large file shares is a significant change that provides an option for highly available file shares with FSLogix. However, in most scenarios, the recommendation is to use Azure Files premium with FSLogix due to IOPS support above 20,000. Currently, Azure Flies premium does not support geo-redundant storage. FSLogix Cloud Cache may be a better option if an FSLogix implementation requires over 20,000 IOPS and replication to a secondary region.
Please get in touch with our team to learn more about Azure Virtual Desktop and FSLogix storage options or discuss any related initiatives!