The present Cambrian explosion of complexity in cloud-native, container and microservices infrastructures, with all its potential benefits, does bring with it one major burden: it siphons potential value from the core of the business by diverting resources to manage the complexity. As infrastructure scales to support the business, ever increasing resources are shifted to supporting the various runtimes, platforms and frameworks that form the substrate, and the opportunity to innovate on the central value-making proposition is lost.
Operations, by whatever name, has been dealing with managing business applications on complex, hybrid and heterogeneous infrastructures since the beginning. It’s never been easy. But the demands to adopt, deploy, manage, and scale emerging technologies in production, across a number of clouds, and connect them back to legacy systems and data sources, are now coming faster and more furiously than ever before. “Digital Transformation,” arguably, puts the most pressure on Operations. And of course, there remains the tacit expectation that Ops, as a cost-center, remain as lithe, agile, cross-functional and responsive as ever.
The answer is found, as it always has been, in use of force-multiplying technologies, and there are plenty to choose from. Most approach the problem from the edges, for example, templates to deploy groups of VMs, domain-specific tools to manage configurations en masse, competing container and image formats, application performance management and monitoring, etc.
All address portions of the overall problem, in some cases extremely well, but they miss the essential requirement: Businesses want the value-creators — their applications — to run in a way that maximizes their value and minimizes operational cost and risk. The solution is not finding the perfect combination of runtime, scheduler, packager, configuration management, telemetry, deployment tool, integration suite, etc. The solution lies in a control plane that works with these, treating them as interchangeable units, unifying them in support of the ultimate goal.
Cloudsoft AMP is the only solution to take a top-down approach, and it is the only management approach which embraces all the right tools for any job.
Cloudsoft AMP has the unique value proposition that it is the only application management approach that can address all the complexities, and create a unified whole from the disparate infrastructure support elements that support a business. At its heart, AMP is a control plane that focuses on policy-based operations and autonomic in-life management; it’s a framework that allows a business to model applications as discrete units of high-level functionality and abstracts away dependence upon the implementation details of any specific infrastructure. The challenge of operations is addressed in a three-part approach: Modeling, Deployment, and in-life Management.
An “application” in the view of AMP is any arbitrarily complex deployable technology: A simple three-tier web application, Kubernetes, SAP, a Docker Swarm, or a Spark-based analytics solution are some examples. Application components and their (dynamically set) configurations are packed into flexible, reusable and composable blueprints. Complex application deployments are built by linking these self-contained, self-managing and self-optimizing entities together. A declarative, YAML-based syntax is used, and the blueprints are therefore extremely readable and easily modified.
A model isn’t much use if it can only be deployed to specific infrastructure, so with AMP, blueprints are designed to be deployable, with little to no modification, to all major cloud providers, to physical infrastructure nodes, to container management runtimes like Kubernetes and Docker Swarm, and, critically, to combinations of these locations. All enterprises are, or will become over time, hybrid, and infrastructure will become fully fungible; applications need to be deployed and managed in a way that takes advantage of these tidal shifts.
The value that Cloudsoft AMP brings is its robust and mature in-life management. Performance requirements are expressed as policies which trigger the appropriate actions automatically. HA, data sovereignty, security hardening, emergency response. Anything that might normally be controlled or implemented by a person or a process is applied as an inheritable, composable policy definition.
Compliant behavior of an application in the largest scope emerges as a natural consequence of these policies autonomically managing application entities. In this way, there is no separation between the development of an application and its deployment or operations. And there is no opportunity for a runbook or a process to be out of date: All are captured and codified within the blueprint model.
There is a good deal of overlap in the tools and approaches in the general space of application and infrastructure management. However it’s wrong to believe that they are all solving the same problem in different ways; Rather, each attempts to solve a small slice of the pie, and there is technical encroachment onto each other’s slices. This is why our general approach has been to integrate the important technologies in support of our primary goal. For example, Docker Swarm and Kubernetes are both supported, as is incorporating Chef recipes within AMP blueprints. Because Cloudsoft AMP is not attempting to be the runtime platform, it can be agnostic about where and how applications run, letting businesses use the best technology for the job in both green-field, and in brown-field and allowing businesses to introduce consistent cloud-native modeling and management without any re-architecting.
Cloudsoft AMP is the only solution to take a top-down approach, and it is the only management approach which embraces all the right tools for any job. A litmus test when comparing possible solutions is whether it can encompass the entire breadth of the systems that need management, automation, and compliance: if a tool doesn’t talk to services, or it doesn’t support legacy, or it doesn’t control wide-area networking, then it isn’t the control plane the business needs. These tools can be useful ingredients in a modern application management solution, and AMP will work with them, but Cloudsoft AMP is the only such universal control plane.
Cloudsoft is a sponsor of The New Stack.
Feature image via Pixabay.
The New Stack is a wholly owned subsidiary of Insight Partners. TNS owner Insight Partners is an investor in the following companies: Velocity, Docker.