Skip to main content Skip to Footer

BLOG


March 15, 2017
AWS’s Snowball Edge speeds data migration to cloud for enterprise
By: Kyle Smith

Migrating to AWS? You are probably going to need your data! And if you are like most large enterprises, there is a lot to move. Don’t count on the network alone to move it. When migrating terabytes upon terabytes of data, plan to use AWS Snowball Edge.

The alternative, transferring that data over the network, is prohibitive due to bandwidth constraints, even when using parallelization. Over the network, for example, it could take months for an enterprise to migrate application data, databases, data warehouses, archived and point-in-time data snapshots—even for traditional lift and shift migrations.

Announced at re:Invent 2016, AWS Snowball Edge is a physical device with 100 terabytes (TB) of storage and built-in compute capabilities. Approximately the size of a carry-on suitcase, the physical drive is a necessity for companies planning to transfer large datasets quickly and securely into the AWS cloud, either in a one-time mass migration or as periodic rolling updates.

Companies with regular archive cycles or frequently updated large data sets (i.e., global sales transactions, mobile phone call records, social media data), can also use AWS Snowball Edge in a round-robin capacity, swapping full drives for empty ones. This ongoing data migration, with the addition of automation to reduce human error and decrease time requirements, makes it possible to regularly add to the data store already in AWS cloud.

New compute capability and other features

AWS Snowball Edge (which builds on the predecessor AWS Snowball service) now has compute capabilities embedded into the hardware device. On-board compute enables a larger set of use cases and can reduce the number of network hops required to get data on the device. With local compute, AWS Snowball Edge devices can be distributed to more remote places to capture data closer to the source of origination, such as collecting telemetry data on a boat, train or even at a remote factory. This is a great way to reduce network bandwidth consumption when having to centralize large amounts of data.

AWS Snowball Edge can also run a small subset of AWS services, like S3 API endpoints and limited Lambda functions (currently supporting Python) on the device, in order to perform functions like transcoding, encryption/decryption, compression and other data transformation functions.

The drive has a multi-custodial nature that is fundamental to the process. A typical migration will see it pass from client to courier to AWS admin. This may not be a pattern that enterprises in highly regulated industries, like financial services and healthcare, are used to when handling their most sensitive data. Therefore, Accenture recommends giving prioritization to the AWS Snowball Edge service when developing a service adoption road map, to provide enough time for Information Security and Audit approvals.

AWS Snowball Edge devices provide AES-256 encryption that does not require storing an encryption key on the physical drive. Instead, customers use the AWS Key Management service to store the key used to encrypt the data before migrating to AWS. Customers then access the key to decrypt the data after it has been migrated to S3 in the chosen AWS region.

Getting started with AWS Snowball Edge

To use the AWS Snowball Edge service, companies submit a request to receive a device using the AWS Console or Command Line Interface. The cost for 10 days of use is a flat $300. Each additional day of use is priced at $30/day, and standard S3 data storage and transfers out costs apply.

Efficient use of AWS Snowball Edge requires proper planning of how the device will be connected to the data source at the customer site. In addition to needing the approval of Information Security, it is important to understand how quickly the target data set can be moved to the AWS Snowball Edge. The speed of the migration will be constrained by the slowest point in the network topology from the source of the data to the AWS Snowball Edge device.

Note: AWS Snowball Edge can be used to migrate data into or out of AWS cloud, but it cannot be used to move data across AWS regions. Therefore, companies using this service must perform a multi-step process to move data from one AWS region to another.

By packing local compute and massive storage capacity into a low-cost, highly portable device, AWS’s Snowball Edge makes it faster and easier for companies of all types to make the journey to cloud.

Popular Tags

    Archive