Optimizing for Cloud- Capitalizing on Your Cloud Estate

Now that the cloud adoption is in full swing, many companies are pondering how best to manage their IT assets. Once a company has overcome the complexities of a migration, optimizing IT applications and minimizing costs should be a priority. The obstacles may seem steep—extensive data to manage at high speeds, lack of control over every element and costs that can escalate quickly—but they are well worth the investment.

While a certain sense of apprehension is understandable when it comes to cloud management, there are eight essential rules that companies can observe to get the most out of their cloud estate.

1. Know your estate

Moving to cloud does not mean turning away from operations once the migration is complete. In fact, once you’ve migrated, it becomes increasingly important to keep an eye on the ‘nuts and bolts’ of your IT function. The first step toward simplifying cloud management and controlling costs is having a thorough understanding of what’s in your IT estate. Are there duplicative or dormant apps? What are the costs? Who has access?

2. Offer equal access to data

While most CIOs and CTOs are reluctant to democratize data, doing so creates a sense of ownership and responsibility across the enterprise. When application owners or individuals from different parts of the business feel accountable for certain data sets, they’re liable to exercise greater caution over spending. However, too much data can be dangerous. Even the most technically and financially astute team members would find a monthly bill with more than 200 million line items overwhelming.

Creating meta tags to parse out pieces of data—who’s responsible for what—helps a company add value while easily tracking information. Typically, up to 10 tags can help simplify the process and better monitor spending and allocation.

3. Establish regulators and set limits

If an organization has thousands of virtual machines constantly changing, it’s usually beneficial to schedule and track what’s being spun up and when. Doing so helps scale assets and monitor activities closely and more cost-effectively. A cloud management platform is a way to automate compute activity with rules that recognize tags for different types of machines. For example, a platform can schedule unused machines to automatically shut off in the evenings and switch back on in the mornings to maximize utility while minimizing expense.

Creating rules and setting limitations around usage is an important, strategic step toward keeping governance—and costs—under control. This is especially critical in an environment where optimization is a key factor.

4. Size properly

Many organizations don’t appropriately estimate just how much cloud real estate they’ll require, which can be a wasteful and potentially costly oversight. When optimizing cloud assets, keep in mind the old carpenter’s maxim: measure twice, cut once. In other words, carefully assessing just how large—or small—your data center must be to accommodate your workloads can spare you grief down the road. Storage volumes raise another potential issue. Because cloud providers charge based on storage provisioning, as opposed to what’s consumed, assessing just how much storage you’ll need at any given time can save costs and avoid potential headaches.

5. No zombies

In cloud terminology, a zombie is a resource that costs money without offering any real value. It could be a server spun up exclusively for testing or, more frequently, a “mystery server” that no one claims ownership for. One way to manage zombies is to simply shut them down for one week. If no one inquires why, chances are it was a zombie.

6. Lose the servers

The least expensive and easiest servers to maintain are virtual. In the server less model, the business using the system is not responsible for purchasing, renting or provisioning the servers or virtual machines upon which code runs. Under this model, resources are transparently allocated at runtime as they become available by the cloud provider. Lamda is one such model—a pure pay-as-you-go enterprise that saves costs and leaves nothing to waste.

7. Don’t stockpile your data

Keeping every single file can lead to wasted resources and big costs. Picking the right storage solution requires a tradeoff between expenditures and functionality. Storing archival data on a costly drive, for example, is clearly a waste of good resources. The cloud offers many versions of storage—from extremely fast (and thus, more expensive) to more cost-effective cold archive storage. Also important, while an organization optimizes the bulk of its IT estate, storage becomes an ever-more critical cost consideration as it proceeds to migrate and mature in the cloud.

8. Keep an eye on performance and costs

Because the cloud is a dynamic and continually evolving entity, it requires rigorous observation—nothing less than constant vigilance. An organization should react quickly if, say, it learns that expenditures have gone over limits or data is compromised. Many companies delegate a full-time resource to monitoring cloud activity.

As cloud adoption becomes mainstream, it will become increasingly critical to have a solid management strategy in place. Establishing the right cost and data management, governance and cloud optimization tools is an intelligent and proactive way to manage your IT estate, helping ensure you maximize the impact of your technology assets.

About the Author: Rodrigo Flores is managing director of Product Innovation, Architecture and Management for Accenture Cloud Platform. The Accenture Cloud Platform is a multi-cloud management platform that procures, provisions, orchestrates, manages and governs enterprise cloud resources.

Edited by Mandi Nowitz