Based on my reporting for a recent deep dive into cloud frameworks for Computerworld and my ongoing messaging work with my clients, here’s what IT customers are looking for when you pitch “grown up” cloud management capabilities to them.
• The scale to accommodate hundreds of applications, and the changes made to them, by thousands of developers. This is a level of complexity far above what many management applications, and even some cloud frameworks, can handle. It may require, depending on the customer, integration with authentication, access control, change management and provisioning systems.
• Visibility into the cloud, in the form of real-time utilization statistics and the ability to assign a monetary cost to, say, a dollar of compute time for a specific type of server or for a week’s worth of storage. This is essential to charging back costs to various departments, or to “show-back” (letting a department or user know what they’ve used.) Without this visibility, organizations can’t decide when they’re spending too much on a given function or whether a public cloud such as Amazon would be a better deal than their internal infrastructure. A related need: Knowing the geographic location of specific servers or databases for security or compliance reasons.
• Service, not just server, management. This can mean a business service, such as email, claims underwriting or security, or more fine-grained services such as the code required to price a loan or generate an invoice. Large organizations looking for major improvements in agility and cost are creating services that can be tracked and reused over time. Since many of these services will live in the cloud, management platforms need to be able to find, track and manage them. Note: Many cloud frameworks claim to provide portability across cloud platforms, but customers say that’s no good without the ability to move supporting services such as database integration and security.
• Life cycle management of systems that goes beyond the up-front provisioning of servers. This includes the ability to provision not only servers, but their associated network and storage resources, and to automatically change the resources allocated to them as needs change. To assure security and compliance, along with efficiency, this requires the ability to create and manage management policies. If, for example, a developer spins up a server for an agile development sprint and then moves on to other work, the management tool will automatically shut down that server if it hasn’t been used in three weeks.
• Policy-based automation in everything from the discovery, inventorying and monitoring of cloud resources to provisioning and reconfiguring the cloud. As clouds become larger, more complex and more critical, doing this work manually becomes so error-prone and complex it’s almost unfeasible even if a customer could afford it.
Is this an impossible shopping list? Maybe so. But even if you can only hit one or two of these targets, focus on the fact you can provide it, explain clearly how you do it and above all prove it with customer testimonials and demos.
Filed under: Uncategorized
Like this post? Subscribe to my RSS feed and get loads more!