One of the benefits of deploying Azure File Sync (AFS) is the ability to cloud-tier your code file server storage –it’s the primary reason, in my opinion. You can enable file sync in a granular manner: for each server endpoint in each sync group. For example, if I had a sync group for a share called Marketing with server endpoints synchronizing this share between three different file servers, I could enable a different tiering policy on each server.
When you enable tiering, the mechanism should not change how users see or use the data, or how permissions are applied – and that’s exactly how AFS does things. When a file is tiered to the cloud, the cold file is removed from the file server, and is replaced with a pointer or reparse point, with the same name and permissions, that points to the cloud copy of the file. From the users’ point of view, little has changed:
Note: the tiering is handled by a filter driver called StorageSync.sys.
As I mentioned before, you can configure tiering on each server endpoint that you want to enable tiering on.
The settings are simple:
When the sync agent on this file server receives the above policies, it will not implement three policies. It will take the largest amount and apply it across the three folders. The coldest files from all three endpoints will be tiered until 30% of drive space is free on the drive – if it is possible to free up 30% of drive space by tiering these folders.
There are some requirements for tiering: