It has been estimated businesses overspend on Amazon Web Services by 35% to 45%, yet there are some simple steps you can take to reduce an AWS bill. The key to these steps being effective and long-lasting is to have total visibility of your assets using a cloud management platform such as CloudHealth.
Nobody really knows for sure how much businesses overspend on Amazon Web Services. Estimates claiming an average 35% to 45% overspend are based on businesses whose AWS bills have been reduced by these percentages after implementing cost reduction best practices. There could be plenty of businesses overspending by more, and some overspending by less.
From the experiences of businesses who’ve managed to reduce an AWS bill, overspending on Amazon Web Services can be attributed to a number of factors. Most often these include overprovisioned assets, assets left running when they aren’t required, assets that should’ve been terminated when they were no longer needed, and a failure to take advantage of Reserved Instances.
Businesses aren’t exclusively at fault for overspending on Amazon Web Services. Cloud Service Providers claim businesses “only pay for what you use”, when the truth is you pay for what you provision. If you provision too much capacity, leave assets running when you aren’t using them, or fail to terminate assets when they’re no longer needed, you’ll still be billed for them.
In order to reduce an AWS bill, one key element is essential - visibility of all your assets. You can’t control what you can’t see, so it’s important to implement a cloud management platform such as CloudHealth that collects data from all sources to give you a single pane view of your infrastructure. Once you have total visibility, you can take these simple steps to reduce an AWS bill.
CloudHealth not only collects and consolidates data into one easy-to-view place; it also evaluates the utilization of each asset and makes rightsizing recommendations to save you time analyzing your assets individually. CloudHealth will also alert you to other ways in which you can reduce an AWS bill - for example, moving infrequently accessed data to a lower cost tier of storage - or ways in which you can improve performance by upgrading assets to a more appropriate configuration.
Assets that aren’t required around-the-clock can be scheduled to start and stop when you want to use them. Some businesses do this already, but apply a universal schedule to all non-production assets - for example, 8:00 a.m. and 8:00 p.m. Monday to Friday. This schedule will reduce an AWS bill for non-production assets by 65%, but further savings can be achieved using CloudHealth’s “Smart Parking” capabilities to schedule start/stop times for when individual assets are actually used.
During the evaluation process, CloudHealth identifies assets with low or zero utilization, and assets surplus to requirements (i.e. aged snapshots). Most business will likely have a significant number of “zombie assets” suitable for termination - such as unused storage volumes, unattached IP addresses, and idle Elastic Load Balancers - because these components are often automatically attached to an EC2 instance when it’s launched, but not automatically terminated when the instance is terminated.
If the past, some businesses were reluctant to take advantage of Reserved Instances because they were locked into a specific instance type. However, businesses now have the option to convert, modify, and exchange Reserved Instances as their needs change. CloudHealth looks for opportunities to purchase Reserved Instances, tells you when they may need converting, modifying, or exchanging, and also alerts you when a Reserved Instance is about to expire so you don’t revert to paying On Demand prices.
Taking these simple steps to reduce an AWS bill will result in instant results; however, it’s important to maintain the results. Due to changing demand, changing workloads, and the constantly evolving nature of cloud computing, there will be times when assets need rightsizing, when schedules need adjusting, and when your infrastructure starts accruing zombie assets. Therefore, to prevent unnecessary costs being added to your AWS bill, your infrastructure needs to be constantly monitored.
CloudHealth monitors your infrastructure around-the-clock using policy driven automation. You simply tell the CloudHealth platform what policies you want to apply and the action(s) you want the platform to perform when a violation of a policy occurs. For example, you could configure CloudHealth to alert you when month-to-date spend is projected to exceed the monthly budget, or when the utilization of an asset falls below a user-specified percentage.
Love the cloud but not loving your cloud spend? Download the 5 Best Practices for Improving Cloud Cost Management eBook.