Last Update: Sep 04, 2024 | Published: Sep 17, 2018
I explained how the components of Azure File Sync interconnect in The Architecture of Azure File Sync. When you deploy a sync agent to a file server, each replicated folder, referred to as a server endpoint, synchronizes to an Azure Files share, referred to as a cloud endpoint. This combination us referred to as a sync group. It is possible to have more than on server participate in a sync group. This allows us to synchronize folders between file servers, using Azure as the middle-man.
The file servers do not have to synchronize all folders to each other. For example, let’s pretend there are 3 file servers:
Chi-FS1 synchronizes:
Dub-FS1 synchronizes:
Syd-FS1 synchronizes:
The process of adding a file server to an existing sync group re-does what is done to register the first file server. You will start by ensuring that the file server meets the requirements of Azure File Sync. Next you will:
The sync agent will need to be registered. This will require you to sign into your Azure Subscription in the registration wizard on the file server. Note that Cloud Solution Provider (CSP) customers must toggle a switch to say that they are using a CSP subscription, and copy/paste their tenant ID, which they can obtain from the Azure Portal (Azure AD > Properties > Directory ID). During the registration, you will select the subscription, resource group, and the sync service that contains the sync groups that this file server should join.
You should repeat the above process for each additional file server that you wish to register with this deployment of Azure File Sync. Note that there is an agent charge for each additional file server that you register with the sync service.
Once a server is registered, you can add it to existing synch groups. Open the sync service in the Azure Portal and browse to Sync Groups. For each sync group that you want to extend to the new file server, do the following:
A few minutes later, the folder will start to synchronize on the new file server. Even without tiering, all of the files will appear as reparse points or folders that redirect users to the cloud copies of the files, with the expected paths, names, and permissions; this allows users to start using the file server within a few minutes of being added to the sync group(s).