Cloud storage may seem overwhelming at first, and Amazon Web Services has a number of offerings. IT pro Rick Vanover explains the differences between these storage technologies.
When it comes to cloud storage, there are plenty of products to choose from. Specifically from Amazon Web Services, there are two main storage offerings: EBS and S3, depending on how they are referenced (see my post earlier this week on AWS’s new server side encryption offering).
Let’s first start with EBS. EBS is the Elastic Block Store cloud. This means that it is a traditional “block” storage resource that is provisioned to other AWS services. This doesn’t mean that we can provision iSCSI targets over the Internet to local resources in our data center (and nobody should want to do that!), but it does mean that dedicated resources within AWS can access this storage. So, let’s now focus on the Elastic Compute Cluster, or EC2. EC2 is one among the core services running in AWS, and basically, EC2 provides virtual machines in the AWS clouds. We don’t get to manage the underlying hosts (based on the Xen hypervisor) as EC2 consumers, but we do get to use the workloads running there (without ‘administering’ that storage). These virtual machines are known as Amazon Machine Images (AMIs). The individual AMI is provisioned as an explicit EBS resource, which is in an underlying iSCSI disk resource that is provisioned to the hosts providing the AMI within AWS.
The other type of storage popularly associated with AWS is the Simple Storage Service or S3 cloud. This storage resource is very commonly used with applications that leverage the Internet for content delivery of things like images, videos, etc. Technically speaking, S3 storage resources are web services, meaning, they are accessed through an HTTP interface such as REST. I use a few products, such as TwinStrata, Nasuni or CloudBerry to back up to the S3 cloud via applications to move data, and a number of applications will upload or exchange directly with this service. I always take great interest in finding out how web-based application companies architect their services around cloud technologies such as Amazon. I wrote a blog about how ShareFile avoided disaster with the AWS outage earlier this year.
Outside of the Amazon clouds, other cloud storage resources exist mostly like S3 in that they are web-based. Of course hybrid solutions, such as the TwinStrata CloudArray and Nasuni Storage Controller can present a local iSCSI target on a cache, with the backend of the file network provided on a web-based cloud such as S3.
From the beginning, it is important to know how these two technologies work at a high level. From there, individual cloud solutions can be selected to meet individual requirements. What do the two main types of cloud storage mean to you? Share your comments below.