AWS upgrades in 2016 were an evolution -- not a revolution

AWS upgrades in 2016 were big on enterprise demands with a range of services to aid in migrations, and a host of additions aimed to fill in gaps and expand on existing tools.

For AWS, 2016 was about making useful, if not radical, changes to its platform.

That may seem an odd description for a company that released so many new services and AWS upgrades in the past year, but it reflects the maturing of a platform that continues to dominate the market. There was less attention on radically new services and more focus on gaps in the platform and enterprise demands.

There were no flashy releases as in the past several years, such as Lambda or CodePipeline. Rather, Amazon built on those services and others as it continued to add AWS upgrades from what was once just infrastructure as a service.

"We're really beginning to see the maturation of Amazon as a platform instead of just a hosting environment," said Jason McKay, senior vice president and CTO of Logicworks, an AWS managed service provider.

AWS Lambda, which runs code without the need for users to provision servers, added support for multiple languages. It can now be embedded into devices via the IoT-centric Greengrass. It's also included in the latest version of AWS' physical device Snowball, which transfers large amounts of on-premises data onto AWS. The ability to use multiple Lambda functions in conjunction also improved with Step Functions.

Other notable additions for newer applications included: CodeBuild to test and integrate code for continuous delivery; machine learning services for speech and facial recognition; Application Load Balancing for Level 7 content-based routing; and new instance types for GPUs and field-programmable gate arrays.

AWS also sought to expand data analytics and improve integration of its various services. New capabilities for Aurora and Amazon Relational Database Service improved replication, security, hooks to Lambda and S3 imports. The company added new analytics services for business intelligence, mobile devices, batch processing and SQL queries against S3. Glue, a new fully managed ETL service unveiled at re:Invent earlier this month, will help tie together disparate data sources on AWS.

Data analytics is a particular area of focus for Amazon, as the Aurora managed database continues to be the fastest-growing service on the platform.

"This is an area we're seeing a lot of movement on a customer front," said Varun Dube, global practice head of AWS at Wipro, an IT consultancy. "We have done some pretty large customer engagements and they're not only moving infrastructure, they're consolidating business analytics."

All about the enterprise

Amazon centered many of its AWS upgrades this year on enterprise migration. There are two cases in point. First is the partnership with VMware that will eventually carve out space to run vSphere environments on AWS. Second, is the massive Snowmobile truck that can transfer exabytes of data from customers' data centers.

Other new ways to get data more easily into AWS include Server Migration Service for automated replication of VMs to the cloud, S3 Transfer Acceleration for faster migration via its edge network, and several AWS upgrades in capacity and capabilities on the aforementioned Snowball. Application Discovery Service identifies on-premises application dependencies to prepare for the transfer to AWS.

"[AWS is] doing a lot of work to educate and reach out to [the enterprise] community to really bridge the gap," said Patrick McClory, senior vice president of platform engineering and delivery services at Datapipe, an AWS managed service provider.

Some new features in 2016 are intended to improve application management. AWS Managed Services is tailored to traditional enterprise workloads ported to the platform. X-Ray analyzes and debugs applications in production, while Shield protects against distributed denial-of-services attacks. The general availability of Amazon Inspector helps automate security assessments. Application EC2 Systems Manager, a late-year addition that flew under the radar, has big potential for security and automation that frees workloads from specific servers.

AWS added four new regions to get closer to customers. They include Canada, where there has been considerable demand for expansion north of the U.S. border. Across regions the total number of AWS instance families grew to 11. Amazon Lightsail provides a simplified way to run a virtual private server and Glacier adds faster retrieval times for cold storage.

More of the same in 2017

With more than 90 different services, plus challenges around pricing, many customers remain concerned about the platform's complexity. Overall, Amazon is expected to continue to fill gaps and build on the platform's tool sets in 2017.

Amazon remains far ahead of the competition with the depth and breadth of its services, but a few areas should be fine-tuned, such as the cost of high IOPs constructs compared to the competition and size changes to attached EBS volumes, Dube said.

"AWS has its own strategy on maturing its products with time, so if you look at the fine print on services there are gaps that need to be filled," he said.

Trevor Jones is a news writer with SearchCloudComputing and SearchAWS. Contact him at [email protected].

Next Steps

VMware, Amazon move forward with partnership

AWS Managed Services targets larger enterprises

AWS re:Invent 2016 unveils new services

Dig Deeper on AWS industries and vertical markets