Cloud services can be set up quickly and flexibly, but with each new application, the complexity also increases. This means that the overview of the number and price models of the various resources can be lost quickly. How can users be informed about rising costs? Which services and projects are responsible for this? Who are the appropriate contacts? This article provides an overview of how Amazon Web Services (AWS) can easily answer these questions and what measures you can take to control costs.

Cloud services are easy to create, but the expected expenses aren’t always clear at the beginning. Concerns about sudden jumps in costs often leads to people reaching for conventional, dedicated server architectures. But in principle, it takes just three planning steps to negate this:

• Pre-estimate costs based upon the planned infrastructure and its characteristics
• Monitor and communicate dynamic cost changes
• Cost identification

Cost accounting

Identifying required services and the initial estimate of foreseeable costs are important steps for delivering infrastructure in the cloud. The AWS Cost Calculator provides an easy-to-use web interface that can quickly and easily calculate expenses for each service based on its required characteristics. We can quickly achieve a high level of cost-effectiveness by minimizing requirements. The possibility of grouping individual cost instances will make hotspots in the planned architecture visible.

Budgets

Initial estimates are important and necessary. However, the dynamic nature of cloud deliveries means that costs will vary, especially when you experiment with services. In the AWS Cost Management service, Budgets provide the first and easiest instance against unexpected expenses. You can create multiple budgets to monitor different time periods and trigger alerts (via email, in the easiest case) when it exceeds a specific amount. To react quickly and automatically, more experienced users can also define actions to restrict identity rights or stop virtual machines and managed database instances with IAM rules. As shown in Figure 1, AWS provides additional forecasted costs for periods greater than daily, allowing us to respond early to upcoming issues.

Fig. 1: Two defined budgets in AWS. Alerts have already been sent for the monthly limit and a forecast for the current month is ready.

Cost Explorer

Once the budget has identified sudden cost jumps or high forecasts, the task now is finding the culprit among the multiple services...