There are couple of ways provided to access S3 Storage On Premesis. To access S3 Storage on-premises you need to configure Storage Gateways.There are 2 ways to do so
Gateway-Cached volumes: Maintain local, low-latency access to your most recently accessed data while storing all your data in Amazon S3.
You will need a host in your datacenter to deploy the gateway virtual machine (VM). Pick a host that meets these minimum requirements.
This support to virtualization technology Windows Hyper-V and VMware.
You have to download the Virtual machine provided by Amazon AWS which is created for this purpose only.You need to import the virtual machine on Hyper-V or VMWare ESX.
Your gateway prepares and buffers your application data for upload to AWS by temporarily storing this data on disks referred to as upload buffer.
Using your Hyper-V Manager client, allocate one or more local disks to your gateway VM for your gateway’s upload buffer. To estimate the amount of upload buffer your gateway requires, use the approximate data you plan to write to your storage gateway’s volumes on a daily basis. It is strongly recommended that you allocate at least 150 GBs of upload buffer.
Again using your Hyper-V Manager client, allocate one or more local disks to your gateway VM to cache recently accessed data on-premesis. These disks, called cache storage, are used to provide low-latency access to data you actively access.
Gateway-Stored Volumes: Schedule off-site backups to Amazon S3 for your on-premises data.