Azure File Sync: solution overview

The Azure File Sync service (AFS) allows you to centralize the network folders of your infrastructure in Azure Files, allowing you to maintain the typical characteristics of a file server on-premises, in terms of performance, compatibility and flexibility and at the same time to benefit from the potential offered by cloud. This article describes the main features of the Azure File Sync service and the procedures to be followed to deploy it.

Figure 1 – Overview of Azure File Sync

Azure File Sync is able to transform Windows Server in a "cache" for quick access to content on a given Azure file share. Local access to data can occur with any protocol available in Windows Server, such as SMB, NFS, and FTPS. You have the possibility to have multiple "cache" servers in different geographic locations.

These are the main features of Azure File Sync:

  • Multi-site sync: you have the option to sync between different sites, allowing write access to the same data between different Windows Servers and Azure Files.
  • Cloud tiering: are maintained locally only recently accessed data.
  • Integration with Azure backup: becomes invalid the need to back up data on premises. You can get content protection through Azure Backup.
  • Disaster recovery: you have the option to immediately restore metadata files and retrieve only the data you need, for faster service reactivation in Disaster Recovery scenarios.
  • Direct access to the cloud: is allowed to directly access content on the File Share from other Azure resources (IaaS and PaaS).

 

Requirements

In order to deploy Azure File Sync, you need the following requirements:

A Azure Storage Account, with a file share configured on Azure Files, in the same region where you want to deploy the AFS service. To create a storage account, you can follow the article Create a storage account, while the file share creation process is shown in this document.

A Windows Server system running Windows Server 2012 R2 or later, who must have:

  • PowerShell 5.1, which is included by default since Windows Server 2016.
  • PowerShell Modules AzureRM.
  • Azure File Sync agent. The setup of the agent can be downloaded at this link. If you intend to use AFS clustered environment, you should install the agent on all nodes in the cluster. In this regard Windows Server Failover Clustering is supported by Azure Sync Files of deployment type “File Server for general use”. The Failover Cluster environment is not supported on “Scale-Out File Server for application data” (SOFS) or on Clustered Shared Volumes (CSVS).
  • You should keep the option "Internet Explorer Enhanced Security Configuration" disabled for Administrators and for Users.

 

Concepts and service configuration

After confirming the presence of these requirements the Azure File Sync activation requires to proceed with the creation of the service Storage Sync:

Figure 2 – Creating Storage Sync service

This is the top-level resource for Azure File Sync, which acts as a container for the synchronization relationships between different storage accounts and multiple Sync Group. The Sync Group defines the synchronization topology for a set of files. The endpoints that are located within the same Sync Group are kept in sync with each other.

Figure 3 – Creating Sync Group

At this point you can proceed with server registration by starting the agent Azure File Sync.

Figure 4 – Initiation of the process of Sign-in

Figure 5 – Selection of server registration parameters

Figure 6 – Confirmation of registration of the agent

After the registration the server will also appear in the "Registered servers" section of the Azure portal:

Figure 7 – Registered servers into Storage Sync service

At the end of the server registration is appropriate to insert a Server Endpoints within the Sync Group, which integrates a volume or a specific folder, with a Registered Server, creating a location for the synchronization.

Figure 8 – Adding a Server Endpoint

Adding a Server Endpoint you can enable Cloud tiering that preserves, locally on the Windows Server cache, most frequently accessed files, while all the remaining files are saved in Azure on the basis of specific policies that can be configured. More information about Cloud Tiering capabilities can be found in the Microsoft's official documentation. In this regard, it is appropriate to specify that there's no support between Azure File Sync with enabled cloud tiering, and data deduplication. If you want to enable Windows Server Data Deduplication, cloud tiering capabilities must be maintained disabled.

After adding one or more Server Endpoint you can check the status of the Sync Group:

Figure 9 – Status of Sync Group

 

To achieve successful Azure File Sync deployment you should also carefully check compatibility with antivirus and backup solutions that are used.

Azure File Sync and DFS Replication (DFS-R) are two data replication solutions and can also operate in side-by-side as long as these conditions are met:

  1. Azure File Sync cloud tiering must be disabled on volumes with DFS-R replicated folders.
  2. The Server endpoints should not be configured on DFS-R read-only folders.

Azure File Sync can be a great substitute for DFS-R and for the migration you can follow the instructions in this document. There are still some specific scenarios that might require the simultaneous use of both replication solutions:

  • Not all on-premises servers that require a copy of the files can be connected to the Internet.
  • When the branch servers consolidate data in a single hub server, on which is then used Azure File Sync.
  • During the migration phase of deployment of DFS-R to Azure File Sync.

Conclusions

Azure File Sync is a solution that extends the classic file servers deployed on-premises with new features for content synchronization, using the potential of Microsoft public cloud in terms of scalability and flexibility.