Select your cookie preferences

We use essential cookies and similar tools that are necessary to provide our site and services. We use performance cookies to collect anonymous statistics, so we can understand how customers use our site and make improvements. Essential cookies cannot be deactivated, but you can choose “Customize” or “Decline” to decline performance cookies.

If you agree, AWS and approved third parties will also use cookies to provide useful site features, remember your preferences, and display relevant content, including relevant advertising. To accept or decline all non-essential cookies, choose “Accept” or “Decline.” To make more detailed choices, choose “Customize.”

AWS Logo
Menu

EBS IOPS: Are You Paying for Performance You Can't Use?

How can you effectively utilize your EBS performance | The Keys to AWS Optimization | S11 E6

cost-savingcost-optimizationiopsebs
Steph Gooch
Amazon Employee
Published Nov 21, 2024
Comments

EBS IOPS: Are You Paying for Performance You Can't Use?

Here are 5 key takeaways from the episodes discussion on EBS performance optimization:

Understanding IOPS Misalignment

Identifying the Problem

Impact on Performance and Cost

Solution Approaches

Best Practices

EBS IOPS: Are You Paying for Performance You Can't Use?

Ever feel like you're buying a sports car engine for a bicycle? That's what's happening when you over-provision IOPS for your EBS volumes. Let's dive into this often-overlooked aspect of AWS optimization that could be costing you more than just pocket change.
In a recent Twitch stream of The Keys to AWS Optimization, Steph Gooch and Justin Marks discussed the often-overlooked issue of EBS IOPS misalignment.

Here are 5 key takeaways from the episodes discussion on EBS performance optimization:

Understanding IOPS Misalignment

  • EBS volumes can be provisioned with more IOPS than EC2 instances can support
  • This leads to wasted resources and unnecessary costs
  • Example: A c5.large instance can only support 20,000 IOPS, but an attached volume might be provisioned with 25,000 IOPS

Identifying the Problem

  • Use AWS CLI to get instance type information:
aws ec2 describe-instance-types —query "InstanceTypes[*].[InstanceType,EbsInfo.MaximumIops]" —output json
  • Utilize AWS Cost and Usage Reports (CUR) data in Athena queries to compare provisioned IOPS with instance limits

Impact on Performance and Cost

  • Over-provisioned IOPS can't be utilized, leading to wasted spending
  • In some cases, this misalignment can result in unexpected performance issues
  • Cost impact can be significant: 5,000 unnecessary IOPS at $0.065 per IOPS/month adds up quickly

Solution Approaches

  • Right-size EBS volumes to match EC2 instance capabilities
  • Consider upgrading EC2 instances if higher IOPS are truly needed
  • Implement regular audits to catch and correct misalignments

Best Practices

  • Review default IOPS settings in infrastructure-as-code templates
  • Consider adding guardrails to prevent IOPS over-provisioning (currently no built-in AWS feature for this)
  • Leverage AWS Cost Optimization Hub for broader optimization recommendations
  • This issue highlights the importance of aligning your EBS performance with your EC2 instances' capabilities. By addressing IOPS misalignment, you can optimize both performance and cost in your AWS environment.
To learn more about this topic and see detailed queries in action, watch the on-demand replay of "How can you effectively utilize your EBS performance | The Keys to AWS Optimization | S11 E6" on YouTube.
 

Any opinions in this post are those of the individual author and may not reflect the opinions of AWS.

Comments

Comments

Log in to comment