Synchronize File Share Content to Windows Azure
Learn about some options to synchronize file servers to Windows Azure.
April 18, 2014
Q: Is there any way to synchronize the contents of an on-premises file server to a SMB file share in Windows Azure IaaS?
A: It's possible to create a virtual machine (VM) in Windows Azure IaaS that can offer file services via SMB file shares. You can use various options such as site-to-site VPN, Multiprotocol Label Switching (MPLS), and other Windows Azure connectivity technologies to connect such a VM to on-premises networks. A popular request is to synchronize the content of an on-premises file server with a file server running on Windows Azure. It's important to realize that the approach you'd use to synchronize two on-premises file servers doesn't change if the target file server is running on Windows Azure. The only difference is that you might see a higher latency from your on-premises network to Windows Azure.
Although it's possible to write scripts that copy files between different file servers, the traditional method is to use Distributed File System Replication (DFSR)—which provides a very efficient synchronization between servers. However, DFSR only works with closed files; if a file is constantly open, it won't replicate. In addition, if a file is edited in both locations, the last edit will overwrite any previous edits. DFSR works best when one share is the focus of changes and the other shares contain read-only type copies.
Some third-party products are available to help troubleshoot synchronization (e.g., Vision Solutions' Double-Take). If the only requirement is to back up data on Windows Azure, rather than have the data actually available on a share in Windows Azure, you could back up to Windows Azure Backup Vault. Alternatively, you could leverage a device such as StorSimple that acts as a local storage appliances accesible via iSCSI and then automatically backs up the content to Windows Azure and even uses Windows Azure as a storage tier to store less frequently used data.
About the Author
You May Also Like