The last part of our IT transformation series is on show-back. The final—some say the first—component of using any cloud for Infrastructure as a Service (IaaS) is cost. We plan to use a cloud to save on costs, but we have to be able to prove we will save money. Unfortunately, with all the approaches to IT transformation—top-down, migration, no changes to management—the truth is that excessive cloud use could turn into massive costs. Those costs, if not handled properly, will end up hamstringing any IT transformation after it happens. The answer is show-back.
Show-back can be as basic as using a simple tool, such as the reports you receive from each cloud on the funds you have already spent for any given month. Each cloud has these reports. If you are using multiple clouds, you need to somehow combine those reports into one, or you need tools to extract that information and show it to everyone who needs to see it. Controlling costs when you use the cloud is a major undertaking, as it is extremely easy to overextend and blow through any budget.
How is that so? Because what the cloud brings to the table is, in effect, elastic resources for a fee. If you deploy 10,000 instances of an application instead of, say, 1,000, you will use perhaps ten times more resources than expected. A simple typo, for example, can increase your costs. Granted, you hope all those things are reviewed before they happen, but we have all heard the stories of large bills from Amazon and other clouds.
With IT transformation, you are hoping to do several things:
- Curtail costs by offloading parts of IT
- Offload IT for resiliency
- Handle rapid expansion of resources in a timely fashion
- Reduce complexity (actually, move complexity into someone else’s hands)
Given that the #1 reason to use the cloud, I hear today, is about cost, it behooves us to understand those costs in some concrete form, to see those costs on an ongoing basis. However, most cloud costs are controversial at best. Most do not account for everything with respect to the cloud, nor do they know what to look into. For example, how are costs allocated after IT transformation? Do you know:
- Cost of cloud resources
- Cost of bandwidth on your end (not the cloud’s)
- Cost of the people to manage/interact with the cloud for development
- Cost of data protection within the cloud
- Reduction in costs associated with any reduction in physical plant (power and cooling) from going to the cloud?
Now, some clouds, such as Virtustream, will bundle all your costs into one fee, a unit of resource. Everything is included in that unit of resource. It makes cost analysis very simple, leaving the costs or reductions associated with your shop to be handled as well.
Show-back gives you a chance to show various folks the savings or extra costs associated with IT transformation. The end result should save on costs. Yet, if you are not careful, it just will not. Some cloud vendors are improving cost controls by applying cost-based limits to what any organization can do. These limits will trigger a sales call to up-sell you on a new tier of pricing, or a new limit. Even with cost-based limits, show-back becomes very important so that you can plan as you approach your limits. VMware, VMTurbo, and others have built cost-based tools to show how much your resources are costing for various services and clouds. VMware’s is limited to vRealize Automation based clouds. There is a new breed of tools that show-back across costs of clouds coming out now. Show-back assists IT Transformation during and after initial deployments. If you cannot get a handle on your costs IT Transformation may falter.
Have you implemented show-back in some fashion related to your cloud services? Even if this means just using the reports generated by the cloud itself? Or do you use some other tool? How has show-back benefited your IT Transformation?
Share this Article:
Latest posts by Edward Haletky (see all)
- Common Product Security Questions - November 23, 2016
- Sorry Support: Not Getting My Data - November 18, 2016
- Moving to the Future: Strategies for Handling Data Scale - November 14, 2016