We’ve identified a handful of AWS cost-driving culprits and also a few ways to quickly spot them in your cost and usage data. Let’s take a look at how to dig into those costs with the best perspectives possible so you can get the right cost and usage insights quickly and take action towards optimizing AWS spending.
Finding a better way to dig into the data
All of these exercise can be done in numerous ways using cost and usage data accessible to all AWS users. But your mileage can vary depending on how you ingest and analyze this data, and which cloud cost management tools you use to surface usable information and insights.
The sooner you get to identifying AWS cost drivers, the sooner you can generate the data points and insights to kick off the right conversations in your organization to improve AWS cost efficiency.
Consider the right arsenal for the job
To sniff out AWS cost culprits Cloudability-style, we’ll be equipped with recently-added proprietary dimensions:
- Lease Type helps compare On-Demand, Reserved and Spot usage
- Usage Family standardizes 8,000 Usage Type values into 20 categories, like Storage and Data Transfer
- Service Name standardizes product naming conventions for AWS and Azure
You’ll see these custom dimensions at work combined with other Cloudability dimensions and metrics within the solution examples below. Learn more about our dimensions, metrics and other analytics additions in our docs.
Our experience shows that AWS users are much more successful at maintaining cost efficient infrastructures when they can quickly understand where their biggest costs are in a few simple clicks by using the right tools for the job. Let’s walk through a few solutions that you can get started with quickly to suss out some key AWS cost drivers.
Identifying unused Reserved Instance hours quickly
Viewing how well your team is utilizing your Reserved Instance (RI) portfolio is an important step in managing overall cloud costs. Infrastructures may constantly shift to accommodate changes in user traffic and workloads, and this presents the challenge of making sure there is RI coverage available for those shifts to avoid paying full On-Demand rates.
One way to immediately identify unused hours is to quickly generate a report that shows wasted RI hours by specific AWS resources. This report is only possible with partial or no-upfront EC2 RIs currently.
In this report, Usage Quantity represents unused hours, Cost (Total) represents the cost of those unused hours and Cost (Total Blended) represents the full monthly payment of the recurring charges. Having these columns side by side and sorted by specific AWS resources (EC2 in this case) lets you quickly get an idea of how many hours aren’t being used and the cost of unused hours. This is a great kind of report and conversation starter after engineering makes significant changes in AWS resourcing.
Digging into data transfer costs
Any service that sends or receives data to another on AWS will likely incur a data transfer cost. Many services will waive these charges, but some have rates by the GB that can add up if left unchecked. To compound the challenge in allocating costs a bit more, AWS invoices and DBR data can track Data Transfer differently. We’ve talked about how to track data transfer costs in the past if you need a primer.
These charges can be difficult to surface with default billing tools alone. The figure above shows a way to quickly surface data transfer charges by Service Name and Usage Family, including a dimension (Resource ID) that lets you hone in on specific resources driving those costs. Less time spent digging in the Detailed Billing Report means more time to find who owns high-cost data transfers and discuss how to optimize them.
Wrangling storage costs
For larger AWS infrastructures out there, there’s a big chance that you’re using a lot of storage across the different types of services. With this many ways to store data on AWS, tracking their individual costs and additional charges (if applicable) can get complicated.
Using the right cloud cost management tools can help you identify individual storage costs, down to specific resource IDs if necessary, to give your team a sense of how much they’re spending on storing things and where the big dollars are heading. Historically, this kind of exercise helps businesses sometimes identify stored data that they didn’t need anymore, but was eating up budget monthly.
For more on optimizing AWS S3 and Glacier costs and usage, check out our comprehensive eBook.
Many more solutions await
Quickly identifying wasted RI hours, digging into hidden data transfer costs and wrangling expensive storage costs are just a few angles that can help decipher what you’re spending on AWS and how efficiently you’re doing it. These presented solutions are also just a sampling of how a well-tagged cost and usage data source, a strong culture of cloud cost management and the right custom reports and dimensions can unlock all kinds of AWS cost efficiency insights.
Having the right data to show the right people on your team also kicks off all kinds of discussions on how to use AWS better, faster and as cheaply as possible. See this in action by trying Cloudability for free. Or get in touch with our cloud cost experts for a walkthrough of these custom reports and dimensions to start transforming your cost and usage data into actionable insights.