Cloud vendors have been trying for years to convince IT professionals to run as many workloads in the public cloud as possible. There are many compelling reasons to migrate some workloads to cloud computing. However, there are also disadvantages.
Public cloud costs can sometimes be unpredictable. Sometimes, I have been surprised by unexpected bills from public cloud providers after I used more resources than I thought within a month.
These surprise bills have not been overwhelming, and I’m the first to admit it. I don’t use Amazon Web Services (AWS) for production workloads, which helps to control costs. Even so, the simple truth that I have had unexpected expenses while performing lab-based activities in AWS’ free Tier underscores the fact the costs of operating within the public cloud can be wildly unpredictable.
Reserved instances are a way for organizations to cut costs and make them more predictable.
AWS charges the on-demand rate for creating an AWS instance. There are many factors that affect the on-demand rate. The base price per hour charged for operating an instance is determined by the instance size and the operating system. Instances often incur additional costs for storage or network transfers. Here is AWS’s on-demand pricing.
AWS offers tools that can help administrators estimate costs, such as the AWS Price List API which was launched in late 2015. However, a better option is to use reserved instances. A reserved instance cannot be created or deleted like on-demand instances. It must be purchased for a year or three consecutive years. AWS charges a flat rate for the instance in exchange for you agreeing to use it for a certain period of time. The hourly rate is significantly less than the on-demand rate if this rate is broken down to a cost per minute. Here is the pricing for reserved instances.
AWS offers its customers flexibility in how they pay for reserved instances. AWS offers a variety of payment options depending on the type of instance. AWS usually allows customers to pay upfront (without a monthly fee), part of the cost upfront and the balance being rolled into monthly fees, or monthly.
AWS offers a report to customers that allows them to see how well their reserved instances have been used. This report is called the EC2 Reserved Instance Utilization Report. It is accessed by logging into the AWS console, clicking your account name, and selecting the My Billing Dashboard option. Figure 1.
Figure 1: Click on the account name to open the drop-down menu. Next, select the Billing Dashboard option. To access the report, click on the Reports container once the Billing Dashboard opens. The available reports are shown in Figure 2.
[Click on the image to see a larger view.] Figure 2: Organizations can use the EC2 Reserved Instance Utilization Report to determine how efficiently their reserved instances are being utilized. An organization might want to dispose of a reserved instance if it is underutilized. AWS does not allow customers to cancel reserved images. However, unwanted reserved instances can be sold to other AWS subscribers through the Reserved Instance Marketplace (although there may be limitations).
AWS recently added a new report, the RI Coverage Report, to its Cost Explorer. This report is used to determine how many hours an organization’s instance hours can be attributed to a reserved example. This report could be described as
Controlling AWS costs with reserved instances
