Amazon Elastic Block Store is a persistent, high performance and highly available block level storage that can be attached to a running Amazon EC2 instance.

Amazon’s Web Services team recently launched Amazon Elastic Block Store and made it available to all it’s user base. Persistance Storage has been the missing (and previously talked about) piece in Amazon’s cloud computing platform and now with the introduction of EBS, that hole has been filled to provide basic building block for a variety of solutions.

Before Amazon EBS, the storage in Amazon cloud was local to a running instance. If the instance got terminated, all the data stored in that instance was lost. Solutions to safeguard the data from being lost included scheduled backups to Amazon S3, and in high change environments, frequent backup strategies introduced tedious processes.

With the introduction of Amazon EBS, virtual machines on the cloud have become more like traditional traditional physical servers. If an EC2 instance goes down, the storage volume can be made available ‘quickly’ by attaching the volume to another EC2 instance. Thus, Persistent storage can act as the machine’s file system that will contain all the data even after restarts.

The separation of storage from the machine has made it possible to attach the same volume to any other EC2 instance. EBS provides the capability to snapshot the entire volume to S3, which makes it easy to create clones in the same and/or other availability zones. Applications can take advantage of this to provide redundancy across availability zones (making sure one creates a new volume from the snapshot in the respective availability zone).

Now that one can have persistent filesystem (of any type) and raw block level access, certain class of applications that deal with large datasets, use filesystem heavily for processing and need to be performant will be easier to build on top of the platform.