Cloudability logo
Choosing the Right EC2 Instances to Optimize Your Cloud
Cost Optimization

Finding your most expensive AWS resources (and what to do about them)

By Andrew Nhem on August 22, 2016
With a growing cloud, it can be challenging to wrangle all cost and usage data to find out which resources are costing your org the most. The right analytical tools in place helps users keep an eye on high-cost resources and start the conversations around optimizing AWS cost efficiency.

With hundreds or thousands of AWS resources running, it can be challenging to wrangle all cost and usage data to find out which resources are costing your org the most. Learn how having the right analytical tools in place can help users keep an eye on high-cost resources and start the conversations around optimizing AWS usage.

Let’s explore some ways to identify your most expensive AWS resources and what to do to save on those cloud costs.

Finding costly AWS S3 storage buckets

Taking advantage of the different tiers of S3 pricing requires knowing which buckets are driving costs the most. Once those buckets are identified, teams can cross-reference its usage and contents to determine if a different flavor of S3 is required for this storage (or even Glacier). For folks wanting to know more about how AWS bills for S3, see this article.

s3-ex-1

This view represents a default look at S3– we can see various S3 charges with their usage types and descriptions.

Even with substantial tagging in place, most folks using a basic cloud cost management tool might see the overall S3 storage costs and usage by department, teams, or whatever combination of dimensions set up by AWS tagging. This is a great high-level look at cloud storage costs.

But, what if users want to know which specific buckets are driving those costs? What if they see a spike in cost with no additional buckets? How do they dig deeper?

s3-ex-2

The Resource ID dimension shows specific S3 buckets that are driving costs, making it easier to associate storage cost and usage with the teams or projects responsible.
s3-ex-3

Now, we can see the most expensive buckets and drive conversations around optimizing use or finding ways to lower the bill.
s3-ex-4

Looks like torpedo-backup drives a lot of storage spending– that team has some ‘splainin to do!

Having a means of tracking “Resource ID,” like in Cloudability, separates these costs even further, into specific buckets. Now users have the means to see which buckets are costing their teams the most. Once those buckets are identified, it’s much easier to discuss the contents of those buckets and how different storage types (like Infrequent Access, or Glacier) could help users save. See a closer exploration of this at our Knowledge Base.

Identifying expensive, untagged AWS resources

As extensible and flexible as AWS tagging can be, there are resources that remain difficult to tag. Specific EBS volumes and neglected, untagged RDS databases come to mind. The ability to identify expensive, untagged resources (like what we did with S3 above) can prioritize what teams need to focus on to monitor and optimize their AWS costs and usage.

Let’s take a look at an example that shows how RDS users can identify expensive, untagged databases and the usage costs that those resources drive.

rds-ex-1

In this example, we see costs by usage type, but we don’t know which specific databases are associated with these costs. We also see that a lot of these resources are missing the ‘Name’ tag.

Using basic cost and usage monitoring alone, users can see costs attributed to the dimensions that tagging can deliver, e.g. Production Teams, Departments, Projects, Environments, etc. As for untagged items, those can often show up with an unset tag as large chunks of cloud spending.

But at some point, people will start asking about those untagged, expensive resources and how to allocate their cost. There needs to be a way to dig into what’s untagged to be in a better position to identify costs or tag and monitor those resources better.

rds-ex-2

By adding the Resource ID measure to the example above, users can now see the specific resources of the untagged collection that are driving costs the most. Now, Operations and IT teams know where to start focusing efforts on either optimizing those costs and usage, or where they need to improve their tagging policies and identify the “not set” resources.

See our Knowledge Base for more details on how to do identify expensive AWS resources.

Having the right analytics turns cost and usage data into optimization opportunities

The ability to hone in on costly AWS resources gives Operations and IT teams the means to start the discussions and investigations around optimizing their AWS usage. This can lead to leaner environments and cost efficiencies that the business can appreciate.

Finding a cloud cost management tool with the analytical capabilities to turn cost and usage data into actionable insights is key. It’s going beyond “how much does this cloud resource cost?” and into “how do I optimize our performance on the cloud and its costs?”

To see this level of cloud cost analysis in action, feel free to get in touch for a Free Trial of Cloudability.

Being in the know feels great