EMA Blog: Business Intelligence and IT Management

“We want everyday developers (...) to be able to use machine learning much more extensively.” This is Andy Jassy’s mantra targeted at making AWS the company that turns machine learning into a commodity, similar to what the company achieved for IaaS before. Within this context, the following two new offerings stood out of the glut of machine learning and IoT news at Re:Invent 2017.

On October 25, Cisco and Google announced their hybrid cloud partnership, where Google brings the container runtime (Kubernetes), the platform to provide, manage, and consume APIs (Apigee), and of course a wide range of consumable cloud services (visual recogngition, machine learning, text to voice, etc.). Cisco contributes the hyperconverged infrastructure (Hyperflex) with Kubernetes management (Harmony), networking (Nexus 9k), and hybrid cloud management software (CloudCenter) to integrate Google’s public cloud services with the customer’s local data center.

Azure Container Service (now AKS where the K stands for Kubernetes) is now offers managed Kubernetes as Tech Preview. This new service provides single-line install (az aks create –n myCluster –g myResourceGroup), automated upgrades, self-healing, and scaling. Microsoft promises that the Azure control plane for Kubernetes will remain free (AWS charges for management servers), with customers only paying for worker nodes running applications. Like all other major vendors, Microsoft declares Kubernetes the winner of the container orchestrator and scheduler race and stresses its contribution to the open source project (only 37 commits, compared to 114 by IBM, 668 by Red Hat, and 1543 by Google as of October 25, 2017). Microsoft also stresses the importance of its Draft project (acquired through Deis) to make Kubernetes accessible to developers without any container experience. Today, Amazon does not offer managed Kubernetes, but it is expected that there will be an announcement in this regard at re:Invent in November.

One week after the GA of Azure Container Registry and only two months after the availability of Kubernetes on Azure Container Service, Microsoft acquires Deis, the guys who make open source Kubernetes management software (Helm, Steward and Workflow), from PaaS Cloud Provider Engine Yard. The Deis slogan is “making Kubernetes easy to use.” With the Deis acquisition Microsoft obtains talent and technologies to successfully compete in the container arena. Infusing Windows, Visual Studio, Azure and OMS with easy container management capabilities is key for Microsoft to catch up with AWS and stay ahead of Google Cloud Platform.

When it comes to cloud technologies, discussions often get passionate or even heated. It’s all about the “war of the stacks”, where much Cool Aid is dispensed to get customers to buy into the respective cult. This discussion reminds me of the old days of enterprise IT, where everything was about technology instead of business value. You either bought one thing or the other and then you were locked in for a half decade. Dark times.

When it comes to cloud technologies, discussions often get passionate or even heated. It’s all about the “war of the stacks”, where much Cool Aid is dispensed to get customers to buy into the respective cult. This discussion reminds me of the old days of enterprise IT, where everything was about technology instead of business value. You either bought one thing or the other and then you were locked in for a half decade. Dark times.

When I picked “The Journey to the Cloud” as the working title for one of my fall research projects, I triggered some immediate reactions from colleagues and customers, whose opinions I value. And while I typically do not place too much importance on selecting a working title and also did not intend to just warm up an earlier research piece on that same topic, these reactions prompted me to take a minute to think about how we can best target our research for maximum customer benefit. As EMA is conducting this research for our customers – vendors and end users – why not talk to exactly these people to find out which specific issues they want to learn more about? And so we did…