IBM Cloud Orchestrator smooths out differences across multiple clouds


and redeployment. This only gets more challenging when you move beyond a single cloud.

Multicloud projects, including hybrid and public clouds from multiple providers, compound these difficulties because they use different hosting models. Automating deployment through DevOps tools can ease complications but is predominately designed for data center use and lacks the ability to deal with the elasticity and portability of cloud. Additionally, DevOps tools aren’t aimed at end-to-end, full-scale, operational automation; most are deployment-centric, and the cloud needs more than that. That’s where IBM Cloud Orchestrator, a cloud management platform for operations automation, comes in.

IBM Cloud Orchestrator features and functions

IBM’s cloud vision embraces broader collaboration, smart cities, the internet of things and other business goals that are typically seen as future cloud applications — not pieces of the IT architecture. IBM Cloud Orchestrator reflects the need for more complex application lifecycle management to deal with complicated IT and business frameworks. The cloud management platform organizes and manages applications to support business goals, not just deployment rules.

IBM Cloud Orchestrator puts three elements of cloud deployment into templates for order and control: infrastructure services, application platforms and governance. The graphical interface enables users to define the control structures for each template and import third-party elements for them. All of the elements are integrated through IBM’s Business Process Manager (BPM), which ties back to the root business activities.

IBM’s cloud vision embraces broader collaboration, smart cities, the internet of things and other business goals that are typically seen as future cloud applications — not pieces of the IT architecture.

The applications, infrastructure and platform descriptions are highly abstract patterns, which means users can define a deployment in general terms and then describe how that general approach applies to any cloud or private IT platform. Patterns, which are various steps melded together to create a predefined form, can also reference DevOps tools already in use. This way, the administrator can connect islands of DevOps deployment with end-to-end orchestration. You can use IBM Cloud Orchestrator to harmonize differences among orchestration tools the cloud provider offers, as well as the deployment differences between containers and VMs.

IBM’s cloud management platform is event-driven, which makes it well-suited to control multicloud environments — in which conditions in all the clouds and components hosted there are completely asynchronous. Events trigger operations defined by either the users or third-party suppliers, and these operations are analogous to lifecycle processes.

With IBM Cloud Orchestrator, multicloud users can import, deploy and export Topology and Orchestration Specification for Cloud Applications (TOSCA) service templates as patterns for platform and infrastructure elements. Because TOSCA acceptance is growing, it’s a critical resource to incorporate cloud descriptions from multiple providers. TOSCA models include three plans: structural, build and management. Patterns native to the cloud management platform take this general approach and seem to draw strongly on TOSCA lessons.

Get started with IBM Cloud Orchestrator

The first step to use IBM’s cloud management platform is to define the application’s operations lifecycle as a set of business processes with BPM. This outlines the abstract set of steps to complete for deployment and the events that could arise during the operation, such as failure or scaling. The IBM Cloud Orchestrator console uses simple drag-and-drop steps to simplify this process. Consider hierarchical definitions: A step to deploy payroll systems, for example, might decompose into multiple BPM flows — one for each application.

Next, define patterns for each application and its hosting environment. Patterns can describe individual deployments and cluster, pod or group deployments; users can describe a different pattern for each cloud in the multicloud environment. Remember to define the structure — the workflow-linked map of components — as well as the management events, rules and the process descriptions associated with each event, including the deployment requests.

If all of this is done properly, IBM Cloud Orchestrator will manage the entire application lifecycle automatically. It can balance work among multiple clouds, shift between public cloud and data center, back up one facility with another and so forth. In short, IBM’s cloud management platform organizes an environment that deals with multiple clouds.

IBM Cloud Orchestrator can be difficult to learn and use, and organizations that don’t have a strong IBM commitment could experience some challenges. IBM and third-party partners provide learning resources to combat this steep adoption barrier.

Source link