A handful of our AWS experts, trainers and fanatics attended AWS re:Invent 2017 in Las Vegas to learn about the latest product developments and releases, get their hands dirty in workshops and certification bootcamps, and bring their wealth of knowledge back to Global Knowledge classrooms.
We polled our instructors who attended to find out what they’re most looking forward to in AWS and their advice for how to stay current through 2018.
1. What AWS release or product update are you most excited about? Why?
There seems to be consensus among our instructors-choosing one release they’re most excited about is tough! So here’s a roundup of their responses:
- Inter-Region VPC Peering: For customers with a global footprint (even if you’re just doing BU/DR, you’re using two Regions), this makes transmitting sensitive data more secure and easy to configure and maintain. Until now, organizations had to maintain pairs or more of EC2 instances running VPN software and write complex scripts to ensure high availability and self-healing to keep cross-region connections open.
- DynamoDB Global Tables: Like Inter-Region VPC Peering, this lets customers spread their low-latency, high-performance NoSQL function across many Regions. It offers multi-mater capabilities (which also was added to some RDS engines like Aurora). Pairing this with the other important DynamoDB release (backup and restore capabilities), it greatly reduces the operation burden of maintaining a NoSQL database in AWS.
- Amazon Free Real-Time Operating System (RTOS): Building on last year’s foray into Internet of Things (IoT), the RTOS provides a lightweight OS that’s meant to be loaded into microcontrollers that power IoT devices. This could be a game-changer for companies wishing to make IoT products (and let’s face it, just about everyone is) with integrated front and back ends. I expect they’ll begin adding more integrations between Free RTOS and the back-end core over the coming year.
- Aurora Serverless: Most of the workloads that I deal with are analytics applications using structured data in relational databases. However, with these applications, the load tends to be spiky with periodic batch-oriented runs. For example, a typical run may be three to four hours daily or 36 hours weekly. Aurora Serverless should help significantly reduce the expense to operate these types of applications by providing a highly concurrent database where you don’t have to pay for large periods of idle time.
- Aurora Multi-Master: This announcement for zones and Multi-Master support across regions later in 2018 will reduce the effort required to provide applications with immediate disaster recovery capabilities built in.
- ECS for Kubernetes (EKS): For companies already running Kubernetes (K8s) or running multi-cloud environments, EKS will offload operational burdens of managing a K8s cluster to AWS and allow the ability to more easily move containers between compatible cloud systems.
- ECS Fargate: This brings the serverless model available in Lambda and applies it to containers. We can now reap the benefits of container-based deployments by offloading operational burdens of managing servers to AWS. This will free up developers to focus on code rather than infrastructure!
2. What do you see as an essential AWS skill in 2018?
Greg Hunt, Cloud Architect and Data Scientist: “With AWS, one of the biggest challenges is keeping abreast of new announcements and determining how quickly you can incorporate the new functionality into your existing systems. Given that to take full advantage of AWS you must fully embrace automation, having excellent software development skills is critical.”
Rich Morrow, Principal Engineer: “Breaking the rules yet again, I’ll give you two-humility and the ability to learn quickly (and I’m not even joking here). The avalanche of product releases at 2017 re:Invent now takes the complete catalog of services to over 100. That’s 100 products that have maybe 30 to 150 features each. AWS now covers everything from IoT hardware to networking, compute, storage, developer tools, mobile and content delivery. Knowing which pieces you’ll even use can be a huge challenge and getting proper training from qualified pros is now more important than ever.”
Richard Jones, Principal Solutions Architect: “Deploying containers, leveraging serverless models and anything to do with Machine Learning or AI!”
3. What AWS certification or training do you predict will gain importance in the next year? Why?
Hunt: “I believe that the Associate Solutions Architect certification will continue to be the most important AWS certification. In order to be successful on AWS, you need to have an excellent understanding of the core services that are the building blocks for higher level services.”
Morrow: “I still see Big Data (huge blanket term that now covers technologies like EMR/Spark, Redshift, Machine Learning) as maybe the No. 1 thing most organizations currently need to tackle in AWS. Especially with all of the new releases around Machine Learning and even AWS’ first graph database (Neptune), companies are going to need to ramp their folks up like never before. Luckily AWS (and GK, which uses the same AWS-provided materials) has about three to five (depending on how you classify) courses on Big Data, and I have no doubts we’ll be doubling that number over the coming year.”
Jones: “The AWS Solutions Architect certifications and the newer specialty certifications, Big Data and Networking, will continue to be popular and valuable certifications for organizations and professionals building their career.”