Migrating IT to a provider cloud offers the opportunity to reduce operating costs immediately, because clouds are much more flexible than on-premise systems. This is because unused or little-used resources can be quickly adapted, downsized or eliminated. This usually pays for itself in the following month. However, there are also some factors that can drive up costs. This article describes the five most common ones.
Oversizing instances
When migrating applications to the cloud, it is often not known how large the corresponding instances were in the company’s own data center. As a result, those responsible often apply more resources than would be necessary to avoid bottlenecks. In addition, they are not well acquainted with the performance characteristics of cloud instances, especially at the beginning. Accordingly, many cloud solutions are oversized. As a rule, they are rarely downsized later on.
No automatic monitoring of resource usage
Using cloud solutions enables companies to act agile and dynamically, but that’s exactly what can drive costs. Because if resource usage is not permanently monitored, new costs are constantly being incurred. That’s why cloud governance teams need tools that automatically monitor usage across all cloud infrastructures the organization uses. When they find optimization opportunities, they need to talk to line-of-business managers to reduce excess capacity.
Lack of cost transparency
However, resource owners usually do not know the costs of services or they seem very low to them. This can happen if, for example, the cost of an instance is charged per hour. This can lead to them not realizing how high the costs are for resources that are unused for weeks, months or years. In agile development projects with automatic provisioning of resources, it often happens that services that are no longer needed simply continue to run. This is because there is no process that automatically shuts down resources again.
Complicated billing models
At first glance, many pricing models seem simple. Charges per hour used or per GB of storage space per month are easy to understand. However, the devil is in the details: The three leading cloud providers offer virtual machines alone at more than ten thousand different prices. They depend on many factors such as the region in which they are deployed, the software version and much more. Older versions tend to be more expensive than current ones. Even prices for instances with similar CPU power and memory can vary greatly due to other add-on features. So it pays to compare price lists closely with requirements and make sure there is at least one person in the company who knows the details and is always up to date.
Unclear responsibilities
As long as IT is running in the data center, the central IT department is usually responsible for security, compliance and cost optimization. However, it is not uncommon for the company’s individual cloud services to be controlled by specialist departments. As a result, no one feels responsible for cost monitoring and optimization, and spending gets out of hand.
Optimization is not a one-time thing
One advantage of the cloud is its flexibility. It allows developers and users to quickly access IT resources and solve business problems. But that’s what often makes cost control difficult, because it’s continuously provisioned and scaled, but rarely removed. To use cloud solutions cost-effectively, it’s not enough to clean up once a quarter. Instead, you should implement automated solutions to monitor and optimize cloud costs.
We are happy to support you both in finding cost drivers and in migrating to the cloud or switching from one platform to another. Talk to us!