Amazon Web Services (AWS) has recently launched two products for their customers to support its cloud business. Hitherto Amazon had been focussing on the cost-saving aspects of the public cloud as part of its EC2 agenda. But the launching of these two products, Snowball Edge and AWS Greengrass, indicates that Amazon are planning to adapt a hybrid cloud model in certain instances at least. Hybrid computing involves dividing up applications between local and third-party providers and IT providers like IBM, Microsoft, and Oracle are all unanimous in their opinion about the necessity of a hybrid cloud solution as they are fully aware that large corporations are sensitive about issues pertaining to data vulnerability and geographic statutory restrictions on data movement not to mention their reluctance to cast all their eggs in one basket (cloud in this instance).
Snowball Edge, launched in a conference in Las Vegas, is a follow up to the Snowball storage solution launched earlier by Amazon. While Snowball was a storage device which assisted in companies shipping directly to Amazon for upload to the cloud, Snowball Edge has computing functionality in addition to storage capacity.
Amazon’s AWS Greengrass is a small component, powered by Intel chips and software written using Amazon’s Lambda tool that can be integrated into local devices and run on the cloud. Both the Edge and the Greengrass are ‘on-premise’ tools which means they can be installed at the user’s location. The software can be run either locally or in the AWS cloud. AWS are recognizing that the Internet of Things, (IoT) requires some hardware to be situated locally and hence IoT marks a shift from cloud to hybrid cloud computing. Matt Wood, the general manager of product strategy with AWS defined hybrid computing in an interview with Fortune as “the integration of infrastructure in the cloud with infrastructure not in the cloud”. This shows a direct shift in Amazon’s attitude who earlier took umbrage with enterprise IT people by saying that they should leave many technology tasks up to AWS.
The Lambda tool is used to create ‘event-driven services’. Developers can set up actions that are automatically triggered if the software detects certain conditions, including, for example, users clicking on a website, image uploads, or a sensor detecting a pre-set trigger. If a customer’s home alarm senses motion near the front door, the Greengrass device can deliver a text alert to the customer, a security service, and the police – all individually or en masse simultaneously without the homeowner having to do anything.
Amazon, being a newcomer, will not pose a threat in the near future to leaders in the enterprise segment like IBM, Oracle, and Microsoft. But given Amazon’s resources and size, they could close the gap fairly soon if they decide to give it top priority. The company are still betting that the influx of data into AWS will grow and that companies will need to depend on fewer of their own data centers in the long term and instead allocate that work into a more efficient public cloud infrastructure. In keeping with this belief, Amazon launched its latest data migration tool recently known as the ‘Snowmobile’ which is basically a fortified shipping container that can transport up to 100 petabytes of data to an AWS data centre. Transferring the same amount of data over a network connection could take a few months or more.