Amazon's Elastic Block Storage stretches control to individual volumes


Amazon’s Elastic Block Storage (EBS) is now a tad more elastic.

Thanks to a new feature called , it’s possible to change an EBS volume’s size or performance characteristics while still attached to and in use by an EC2 instance.

In , Amazon claims Elastic Volumes reduces the amount of work and planning needed when managing space for EC2 instances. “Instead of a traditional provisioning cycle that can take weeks or months, you can make changes to your storage infrastructure instantaneously, with a simple API call,” Amazon writes.

On-the-fly changes that can be made to EBS volumes include switching EBS volume types and adjusting allocation. Amazon says, as an example, a user could change out a General Purpose SSD type of EBS volume for the to better handle high-volume data workloads. Another example involved reducing the costs for log storage, where freshly collected logs are kept on a General Purpose volume, which is later changed to Throughput Optimized.

also apply. After changes are made to a volume, at least six hours must pass before further changes can be made. Therefore, any automated changes—for instance, based on performance statistics—should probably be performed two or three times a day, at most. Also, changes to an EBS volume—like a resize—aren’t automatically applied to any of the file systems on a given volume; those changes must be made from within whichever operating system is using that volume.

Elastic Volumes is different from Amazon’s (EFS), which was released to general availability last summer and abstracts away individual volumes. EFS allows developers to create resizable file systems that can be accessed from multiple EC2 instances via NFS. Elastic Volumes, on the other hand, is for those who want to deal with individual volumes and control how space is allocated and what properties it has.