Skip to main content
In this guide, we demonstrate how to use AWS DataSync to transfer files from an on-premises NFS server to an S3 bucket in AWS. The objective is to simulate a corporate data center environment and illustrate the seamless migration process using AWS DataSync. The lab setup emulates an on-premises scenario where an application server accesses an NFS server to read and write data. This data is then transferred to an S3 bucket via AWS DataSync. Although the simulation represents an on-premises environment, all components reside in AWS for simplicity. Below is the architecture diagram illustrating the data transfer flow from the corporate data center to AWS S3. The diagram shows an App Server, NFS Server, DataSync Agent, and the AWS DataSync service:
The image is a flow diagram showing data transfer from a corporate data center to AWS S3 using an App Server, NFS Server, DataSync Agent, and AWS DataSync.

​
Environment Overview

An application server in our simulated environment connects to an NFS server running within an AWS VPC. The data destined for migration is stored on the NFS server and will ultimately be copied to an S3 bucket. The following diagram shows the deployed application server and its connectivity with the NFS server:
The image shows an AWS EC2 management console with a list of running instances, displaying details for a selected instance named "appServer."
To confirm connectivity, log in to the application server and run the command below to display mounted file systems: 
df -h
The output indicates that the NFS server is mounted at /mnt/data with the source path 10.11.12.182:/media/data: 
Filesystem      Size  Used Avail Use% Mounted on
/dev             478M     0  478M   0% /dev
/dev/shm         486M     0  486M   0% /dev/shm
/run             486M   416M  485M   1% /run
/sys/fs/cgroup   486M     0  486M   0% /sys/fs/cgroup
/                8.0G   1.8G 6.3G  23% /
/run/user/1000   98M     0   98M   0% /run/user/1000
/media/data      8.0G   1.8G 6.3G  23% /mnt/data
user@ip-10-11-12-118 ~ $
Navigating into /mnt/data reveals a folder called images that contains numerous JPEG files destined for migration to the S3 bucket. For example: 
[ec2-user@ip-10-11-12-118 ~]$ cd /mnt/data
[ec2-user@ip-10-11-12-118 data]$ ls
images
[ec2-user@ip-10-11-12-118 data]$ cd images/
[ec2-user@ip-10-11-12-118 images]$ ls
00001.jpg  00018.jpg  00035.jpg  00052.jpg  00069.jpg  00086.jpg  00103.jpg  00120.jpg  00137.jpg  00154.jpg  00171.jpg  00188.jpg  00189.jpg  00199.jpg  00200.jpg
00002.jpg  00019.jpg  00036.jpg  00053.jpg  00070.jpg  00087.jpg  00104.jpg  00121.jpg  00138.jpg  00155.jpg  00172.jpg  00191.jpg  00200.jpg  00201.jpg
...
Once the files are transferred, you can verify their presence by accessing the S3 bucket (named “data sync - KodeKloud”), where the files appear under an “images” folder.

​
Deploying the DataSync Agent

The next step is to deploy the DataSync agent. While the agent can be deployed on any supported virtualization hypervisor (e.g., VMware, KVM, or Microsoft Hyper-V), this demonstration uses an EC2 instance in AWS.
The image shows an Amazon S3 dashboard with a list of storage buckets, their regions, access settings, and creation dates. The interface includes options for managing buckets and accessing various AWS services.
Refer to the AWS DataSync documentation for detailed information on deploying the agent in different environments, including VMware, EC2, AWS Snowcone, and AWS Outposts.
The image is a screenshot of an AWS documentation page titled "Deploy your AWS DataSync agent," detailing how to deploy the agent on various platforms like VMware, KVM, and Amazon EC2.
For EC2 deployment, perform the following steps:
  1. Retrieve the latest AMI for the DataSync agent using AWS CLI: 
    aws ssm get-parameter --name /aws/service/datasync/ami --region us-east-1
    The output will be similar to: 
    {
  "Parameter": {
    "Name": "/aws/service/datasync/ami",
    "Type": "String",
    "Value": "ami-dd2f74d8a00ff34d",
    "Version": 88,
    "LastModifiedDate": "2023-10-13T19:30:54.383000+00:00",
    "ARN": "arn:aws:ssm:us-east-1:parameter/aws/service/datasync/ami",
    "DataType": "text"
  }
}
You can run this command in AWS CloudShell if your local AWS CLI is not configured.
  1. Launch an EC2 instance using the retrieved AMI. In the launch wizard, ensure that you update:
    • Source Region: us-east-1 (or your respective source region)
    • AMI: Paste the obtained AMI ID.
    • Instance Type: Options like m5.2xlarge, m5.4xlarge, or c5.medium are recommended. For testing, a smaller instance might be sufficient.
    • Key Pair: Select your designated key pair.
    • VPC and Public IP: Choose the appropriate VPC and assign a public IP address.
    • Security Group: For demonstration purposes, allowing all traffic is acceptable. In production, restrict access appropriately.
The image shows a section of the AWS DataSync User Guide, specifically detailing instructions on deploying a DataSync agent as an Amazon EC2 instance. It includes a highlighted URL and steps for configuration.
The image shows a webpage from the AWS DataSync User Guide detailing Amazon EC2 instance requirements, including recommended instance sizes for different tasks. It also outlines agent requirements for DataSync Discovery and management.
The image shows an AWS EC2 instance launch configuration screen, detailing settings for subnet, security group, and instance type. The summary section on the right provides an overview of the instance details, including software image, server type, and storage volume.
After launching the instance, register the DataSync agent through the AWS DataSync service page:
  1. Click on “Agents” then “Create Agent.”
  2. Select the EC2 option (since the agent is deployed on EC2) and leave the public service endpoint enabled.
  3. Retrieve the activation key by refreshing the EC2 instances list and noting the public IP address of the DataSync agent.
  4. Paste the public IP on the DataSync registration page, assign a name (e.g., “data sync agent”), and create the agent.
A successful registration output may look like: 
Successfully retrieved activation key from agent
Activation key
EVH1S-KH48T-0RF59-K77NG-V5MGL
Agent address
18.234.228.236

​
Creating DataSync Locations

After deploying the agent, the next step is to define the source and destination locations where DataSync will copy the files.

​
Source: NFS File Server

Create a source location for the NFS file server by following these steps:
  • Choose “NFS” as the location type.
  • Select the DataSync agent deployed earlier.
  • Specify the NFS server’s private IP address and set the mount path, which in this case is /media/data.
An example of the connection test for the NFS mount is shown below: 
[ec2-user@ip-10-11-12-118 ~]$ cd /mnt/data
[ec2-user@ip-10-11-12-118 data]$ ls
images
[ec2-user@ip-10-11-12-118 data]$ cd images/
[ec2-user@ip-10-11-12-118 images]$ ls
00001.jpg  00018.jpg  00035.jpg  ...  00209.jpg
The image shows an AWS console configuration page for setting up a Network File System (NFS) location, including fields for selecting agents, entering an NFS server IP address, and specifying a mount path. There are options for additional settings and tags, with a "Create location" button at the bottom.

​
Destination: S3 Bucket

Define the destination location for your S3 bucket by following these guidelines:
  • Select your S3 bucket (e.g., “data sync - KodeKloud”).
  • Use the “standard” storage class (or adjust as needed).
  • Specify the bucket folder if required; for this demonstration, the root directory is used.
  • Assign an IAM role for DataSync to access the S3 bucket. If none exists, you can opt for auto-generation of the role and policy.
The image shows an AWS DataSync interface with details of a created location, including the location ID, type, and associated S3 bucket information. A green notification bar indicates successful creation of an IAM access role and policy.

​
Creating and Running a DataSync Task

With both source and destination locations configured, you can now create a DataSync task to perform the file migration:
  1. Click “Create Task” and select the previously configured source (NFS) and destination (S3) locations.
  2. Name the task (e.g., “copy-nfs-to-s3”).
  3. Configure the task options, including copying all files and optionally setting up logging via auto-generated CloudWatch log groups.
  4. Review the configuration and create the task.
  5. Start the task using the default settings.
The image shows an AWS DataSync configuration screen where a task named "copy-nfs-s3" is being set up. It includes options for task execution and data transfer configuration.
Once the task starts, you can monitor its progress, observing metrics such as throughput, file count, and data transferred from the source to destination.
The image shows an AWS DataSync console screen with a task named "copy-nfs-s3" that is currently running. It displays details about the source and destination locations, including a Network File System (NFS) and an Amazon S3 bucket.
When the task progresses from the launching phase to transferring files (as indicated by metrics reaching 1% and beyond), it confirms successful connectivity between the NFS server and S3 bucket. After task completion, verify the migrated files by checking the S3 bucket:
The image shows an Amazon S3 console interface displaying a list of JPEG files with details such as name, type, last modified date, size, and storage class.

​
Conclusion

This guide illustrated the deployment of a DataSync agent on an EC2 instance, configuration of source and destination locations, creation of a DataSync task, and the verification process for transferring files from an NFS server to an S3 bucket using AWS DataSync. The streamlined integration between these services simplifies data migrations from on-premises environments to AWS, making it an excellent solution for modern data management. Happy syncing!

Watch Video