Offline Data Transfer and Edge Computing with the AWS Snow Family
Learn the device selection criteria for Snowcone, Snowball Edge, and Snowmobile, along with edge EC2/Lambda execution and planning for large-scale data migrations.
Snow Family Device Selection
The Snow Family is a service for migrating large volumes of data to AWS using physical devices when network transfer is impractical. Even a dedicated 1 Gbps line would take approximately 100 days to transfer 1 PB, making offline transfer the realistic choice at petabyte scale. Snowcone (8 TB HDD / 14 TB SSD) is the smallest device, weighing just 2.1 kg in a portable form factor. It is ideal for data collection in remote locations with limited bandwidth. Snowball Edge Storage Optimized (80 TB) is best suited for large-scale data migrations, offering faster and more cost-effective transfer than network-based methods for datasets exceeding 10 TB. Snowball Edge Compute Optimized features 104 vCPUs and a GPU, supporting machine learning inference and data preprocessing at the edge. For ultra-large migrations exceeding 100 PB, Snowmobile (100 PB) is available.
Data Transfer Workflow
The data transfer workflow begins by creating a job from the AWS console and receiving the shipped device. Connect the device to your local network and copy data using the Snowball client, OpsHub GUI, or S3-compatible API. Snowball Edge also provides NFS mount points, enabling seamless data transfer from existing backup tools and file copy utilities. Once data copying is complete, ship the device back to AWS, which then imports the data into your designated S3 bucket. Using multiple devices in parallel enables petabyte-scale migrations to be completed in just a few weeks. After transfer is complete, data on the device is erased following NIST 800-88 compliant procedures.
Edge Computing Use Cases
Snowball Edge Compute Optimized can run EC2-compatible instances and Lambda functions at edge locations. A typical use case is running image inspection on a factory production line, storing results locally, and syncing to S3 when connectivity is restored. Computing continues even in environments with intermittent internet connectivity, such as ships or mines. Since it uses S3-compatible APIs for data read/write, applications developed in the cloud can be deployed to the edge as-is. Integration with AWS IoT Greengrass enables combining IoT device data collection with edge processing. Snowcone supports IoT Greengrass and can be permanently deployed at remote sites as an edge device. For a comprehensive understanding of the Snow Family, related books (Amazon) are also a helpful reference.
Hybrid Migration with DataSync
AWS DataSync automates data transfer between on-premises storage and AWS storage services, and a hybrid migration strategy combining Snow Family with DataSync is highly effective. Perform the initial large-scale data migration with Snow Family, then use DataSync for ongoing incremental synchronization over the network. DataSync supports diverse protocols including NFS, SMB, HDFS, and object storage, with transfers to S3, EFS, and FSx for Windows File Server. Data validation during transfer (checksum comparison) ensures data integrity. Bandwidth throttling controls network load during business hours, and scheduled execution automates periodic incremental syncs to keep on-premises and cloud data continuously up to date.
Security and Compliance
Snow Family devices feature enterprise-grade security. Data at rest is protected with 256-bit encryption (AES-256), with encryption keys managed by AWS KMS. A Trusted Platform Module (TPM) detects physical tampering and automatically locks the device if unauthorized access is attempted. GPS tracking monitors shipping status in real time, and E Ink displays are used for automatic shipping label updates. The devices meet HIPAA, PCI DSS, and FedRAMP compliance requirements, making them suitable for data migration in regulated industries. IAM policies provide fine-grained control over job creation and device access permissions.
Snow Family Pricing
Snow Family pricing is based on device usage days. Snowcone costs approximately $60 per day, and Snowball Edge Storage Optimized costs approximately $300 per day. Shipping fees are charged separately and vary by region. Device usage typically includes 10 days, with overage charged at a daily rate. Importing data to S3 is free, but exporting from S3 incurs data transfer charges. Migrating 100 TB with Snowball Edge costs approximately $5,000-$6,000 including device and shipping fees, which is less expensive than the data transfer charges (approximately $9,000) for transferring the same volume over the internet.
Summary
The Snow Family provides large-scale data migration when network transfer is impractical, along with computing in edge environments with limited connectivity. Choose devices based on data volume - Snowcone (8-14 TB), Snowball Edge (80-210 TB), or Snowmobile (100 PB) - and run EC2 or Lambda in offline environments with edge computing. Combining with DataSync delivers a hybrid strategy integrating initial bulk migration with ongoing incremental sync, while AES-256 encryption and TPM tamper detection ensure enterprise-grade security.