I've taken over an S3 File Gateway that was set up by another team, and it seems to be out of date. We want to streamline operations by moving to Azure Files since that's where most of our SMB shares are located. From what I can tell, the previous team isn't utilizing S3 effectively; it's mainly being used for archiving. I'm trying to figure out if there's a clear migration path away from the S3 File Gateway. The documentation mentions that you can't directly migrate data to an FSx file gateway. My plan is to set up a Windows file server with an Azure Data Sync agent and use robocopy or something similar to transfer the locally cached shares from the S3 File Gateway to the new server. However, I'm not entirely sure how the file shares connect with the S3 archive or if we can successfully pull all the data this way.
3 Answers
The S3 File Gateway is designed to proxy S3 as SMB/NFS shares. If you check the Storage Gateway console, you can find the share names and their corresponding S3 locations. While there’s no direct migration path mentioned, you could set up an AWS DataSync job to transfer files from S3 to FSx as part of your migration.
Your S3 File Gateway acts like an interface for a remote S3 bucket, right? If you’re moving to Azure, try to take the data offline during migration for a smoother process. Azure has guides on migrating directly from S3 to Azure Files, which might help you set up the right connections for your SMB users. Just keep in mind that AWS will charge you for data egress while moving your files, so it might be worth discussing costs with your Azure team. Also, while the documentation says you can’t automate moving data to FSx, you could still use AWS DataSync to facilitate the copy from S3 to your target destination directly.
We can't take it offline either; it's actively used for collaboration, and cutting off access isn't an option. I also don’t see support for migrating directly to Azure Files from S3 in the documentation. Can anyone clarify?
If you’re planning to work with Azure Files without using File Sync, consider launching an EC2 instance that can fetch data from S3 via API and write directly to Azure Files. Personally, I’d create a script to copy everything from S3 to the Azure mounted share. It’s a good idea to run this in AWS for quicker access, but watch out for egress costs. Make sure to set up a VPC endpoint for S3 to help manage those costs.
AWS just rolled out Amazon Interconnect for smoother cross-cloud communication. It might be worth waiting for it to stabilize before making any big moves—talk to your Technical Account Manager about potential costs.
Exactly! Microsoft can help you save on migration costs, so take advantage of that. The trick is copying the data over without cutting off live access.