Skip to main content

10 Steps for an Effective Reserved Instances Strategy

Amazon Web Services (AWS) offers three different ways to pay for EC2 Instances: On-Demand, Reserved Instances, and Spot Instances. This article will focus on effective strategies for purchasing Reserved Instances. While most of the major cloud platforms offer pre-pay and reservation discounts on compute instances, I’ll be using the AWS terminology and context. Whether you are using Amazon, Azure, or Google Cloud Platform, these recommended strategies have proven effective for purchasing Reserved Instances. 

This article assumes you know at least a little about Reserved Instances and the pricing options for EC2 that Amazon offers. If not, be sure to check out the AWS documentation and the summary in this excellent blog post.

Amazon makes several services eligible for Reserved Instance purchases: EC2, RDS, Elasticsearch, Elasticache, DynamoDB, and Redshift. I’ll focus on EC2 Instances, but most of the information is universal. Since these strategies are in a loose order of execution, they are adjustable given your context.

To effectively purchase Reserved Instances, there can be a lot to consider (and to do) before putting together your final order. But it’s worth it given the dollar investment involved and the potential savings for your organization. Here are some best practices for purchasing Reserved Instances.

  1. Engage the key players deploying workloads in the cloud
  2. Engage the stakeholders who will approve your purchase
  3. Run resource utilization reports and right-size analysis
  4. Update instance families where possible
  5. Choose payment options
  6. Decide which accounts to make the purchase under
  7. Start conservatively
  8. Give visibility and educate
  9. Actively manage as a process
  10. Engage your AWS Technical Account Manager

Engage the Key Players

Get to know the key folks in your organization who are deploying workloads in the cloud. These could be cloud ops or DevOps team members. Be sure to also identify the technical leads or architects for the products. They are usually the closest to the product architectures and can help inform you on related technology roadmap decisions. Take your Reserved Instance plans to them for validation — you don’t want to find out they’re going serverless next month after you’ve bought a bunch of Reserved Instances for them. The team is the best source for learning scaling (or de-scaling) plans, instance family changes, and other key data points. You definitely want the warm fuzzies from them before pulling the trigger on any purchases on their behalf.

Engage the Stakeholders

Now that you know the technology decision makers, make sure to identify and engage the stakeholders who will approve your purchase: finance, CTO/CIO, your boss, and other key partners. It’s easy to get lost in the technical details of an Reserved Instance purchase, make sure to keep your presentations at the appropriate level and context. In the end, you’re proposing a strategy to save money and control costs that provide some easy ROI data points: don’t over complicate it.

Run Utilization Reports and Right-Size Analysis

Run your resource utilization reports and identify EC2 Instances that can be downsized, terminated, modernized, or put on automated start-stop schedules. Your organization, as a matter of best practice, should be performing right-sizing analysis and adjustments on a regular basis. This will yield cost savings even with On-Demand pricing, and you’ll want to make sure you are caught up with your right sizing so you aren’t overspending on underutilized resources or buying unnecessary reservations. An Amazon Trusted Advisor can help with this to a certain degree, and there’s a healthy market of effective third-party tools that can automate and streamline your right-sizing process. Some established examples are CloudHealth, Cloudability, and CloudCheckr — and there are dozens more to choose from. The best approach is to identify and define your organization’s needs and evaluate a couple with the features you are looking for. Any of them will save you lots of time by helping you avoid very tedious and complicated analysis.

Update Instance Families Where Possible

Check the instance families your product teams are using. It’s important to look for previous generation families in use and determine if it’s possible to replace them with more updated versions. In addition to performance gains and feature availability, the newer families are usually slightly cheaper. Also, it can help to do some analysis to make sure the right instance types are being used for the workload — with considerations for favoring compute, memory, I/O, etc. Detailed utilization analysis can help determine any necessary instance type changes and make your reservation purchases more effective. If you’re experiencing or anticipating a lot of change in instance types, the convertible option might be best. You’ll sacrifice a few discount points, in exchange for reducing the risk of being stuck with unused Reserved Instances.

Choose Payment Options

Decide on payment options for your reserved instances. AWS offers three approaches to paying for Reserved Instances: all upfront, partial upfront, and no upfront. The all upfront option provides the steepest discount, but it’s usually only 1% or 2% better than the partial upfront. No upfront (essentially just reserving an instance for 1 to 3 years) offers the smallest discount, usually about a 5% difference from the other two options. Between payment options, term length, and reservation type, you can range anywhere from 30% to 60% in savings. Use the AWS Cost Explorer Reserved Instance Recommendation tool in your account console to easily tinker with the options and see their impact on discounts.

Decide Which Accounts to Make the Purchase Under

For consolidated accounts (a group of accounts linked to a master payer account), you’ll want to decide which accounts to purchase the reservations. Consolidated accounts offer a great feature: reservations can float between accounts. This means if the account the reservation was purchased in can’t use the reservation, it can float to another account so the reservation won’t go unused! Given this, some organizations like the simplicity of purchasing all the Reserved Instances in the payer account and letting them float where needed. As a personal preference, I like purchasing within the individual linked accounts, so the reservations will get used first where intended (although that can be a cumbersome purchase). This can help with budget reporting and other detailed cost analysis within the accounts. Note: If you make your purchases in the payer account, you can turn off Reserved Instance discount sharing for specific linked accounts.

Start Conservatively

Don’t go for 100% coverage right out of the gate. It might be more comfortable for all involved to make a solid purchase initially — say, 50-70% coverage, depending on your situation. You can also reduce risk by choosing the convertible option for your reservations. And then you can (and should) setup a regular cadence for updating your analysis, buying more, and modifying or converting existing Reserved Instances.

Give Visibility and Educate

Give attention to visibility and education. Success making technology decisions depends on an informed organization. Providing visibility into your cloud performance and cost metrics can help get teams invested in making the right choices. An important part of any cloud optimization strategy also includes evangelism and education. Folks at all levels will be less likely to “buy-in” to complicated decisions that don’t have obvious connections to an organization’s goals. Sure, the information is all over the internet, but it’s a lot more effective to give it structure and shepherd it along in your organization. That’s also a great way to build partnerships invested in the success of your cloud strategies. A solid training platform (like the one hosting this blog) is also super effective for providing curated information to all organizational role types.

Actively Manage Your Reserved Instances as a Process

The Reserved Instance purchase is not the end. Once you’ve got your Reserved Instances in place, remember to build reviewing and modifying them into your process to reduce the risk of them going unused. Over time, you’ll also need to track, prepare and react to Reserved Instances whose terms are expiring.

Leverage AWS Resources

Engage your AWS Technical Account Manager(s). Have your AWS support team provide you with recommendations independent of your own. They’ve got team members dedicated to this craft and can help with all aspects of the analysis and purchase. This is a great sanity check against your own analysis and can help you avoid missing opportunities or making costly mistakes.

Conclusion

In summary, purchasing Reserved Instances can be complicated no matter the size of your organization’s cloud footprint. Start simple, take the time to understand the foundational concepts, and partner with the key stakeholders in your organization. Working with your teams on right-sizing and modernizing activities as an integrated, routine process will make coming up with effective recommendations for Reserved Instance purchases much easier and less stressful.

 

Further reading:

AWS Docs > EC2 Reserved Instance Purchasing Options
AWS > EC2 Reserved Instance Current Pricing
AWS Docs > Reserved Instance Recommendations in Cost Explorer

Cloud Academy Blog > Managing Multiple Accounts on AWS
Cloud Academy Learning Path > AWS Cloud Management Tools

 

Jeff Hyatt

Written by

Jeff Hyatt

Cloud architect, software developer, disc golfer and life-long geek driven to create, automate, and simplify. With decades of experience managing networks, developing software, and managing teams, Jeff has spent the last couple of years helping enterprises migrate to the cloud and integrate governance and optimization into their processes and culture.

Related Posts

Avatar
Stuart Scott
— July 18, 2019

AWS Fundamentals: Understanding Compute, Storage, Database, Networking & Security

If you are just starting out on your journey toward mastering AWS cloud computing, then your first stop should be to understand the AWS fundamentals. This will enable you to get a solid foundation to then expand your knowledge across the entire AWS service catalog.   It can be both d...

Read more
  • AWS
  • Compute
  • Database
  • fundamentals
  • networking
  • Security
  • Storage
Avatar
Adam Hawkins
— July 17, 2019

How to Become a DevOps Engineer

The DevOps Handbook introduces DevOps as a framework for improving the process for converting a business hypothesis into a technology-enabled service that delivers value to the customer. This process is called the value stream. Accelerate finds that applying DevOps principles of flow, f...

Read more
  • AWS
  • AWS Certifications
  • DevOps
  • DevOps Foundation Certification
  • Engineer
  • Kubernetes
Avatar
Stuart Scott
— July 2, 2019

AWS Machine Learning Services

The speed at which machine learning (ML) is evolving within the cloud industry is exponentially growing, and public cloud providers such as AWS are releasing more and more services and feature updates to run in parallel with the trend and demand of this technology within organizations t...

Read more
  • Amazon Machine Learning
  • AWS
  • AWS re:Invent
  • Machine Learning
Avatar
Stuart Scott
— June 27, 2019

AWS Control Tower & VPC Traffic Mirroring

AWS re:Inforce 2019 is a two-day conference for security, identity, and compliance learning and community building. This year's keynote, presented by AWS Vice President and CIO, Stephen Schmidt, announced the general availability of AWS Control Tower and the new VPC Traffic Mirroring fe...

Read more
  • AWS
  • re:Inforce 2019
  • traffic mirroring
  • VPC
Avatar
Stuart Scott
— June 20, 2019

Working with AWS Networking & Amazon VPC

Being able to architect your own isolated segment of AWS is a simple process using VPCs; understanding how to architect its related networking components and connectivity architecture is key to making it a powerful service. Many services within Amazon Web Services (AWS) require you t...

Read more
  • AWS
  • VPC
Avatar
Stuart Scott
— June 19, 2019

AWS Compute Fundamentals Update

AWS is renowned for the rate at which it reinvents, revolutionizes, and meets customer demands and expectations through its continuous cycle of feature and service updates. With hundreds of updates a month, it can be difficult to stay on top of all the changes made available.   Here ...

Read more
  • AWS
Joe Nemer
Joe Nemer
— June 18, 2019

AWS Certification Practice Exam: What to Expect from Test Questions

If you’re building applications on the AWS cloud or looking to get started in cloud computing, certification is a way to build deep knowledge in key services unique to the AWS platform. AWS currently offers 11 certifications that cover major cloud roles including Solutions Architect, De...

Read more
  • AWS
  • AWS Certifications
Avatar
John Chell
— June 13, 2019

AWS Certified Solutions Architect Associate: A Study Guide

The AWS Solutions Architect - Associate Certification (or Sol Arch Associate for short) offers some clear benefits: Increases marketability to employers Provides solid credentials in a growing industry (with projected growth of as much as 70 percent in five years) Market anal...

Read more
  • AWS
  • AWS Certifications
Chris Gambino and Joe Niemiec
Chris Gambino and Joe Niemiec
— June 11, 2019

Moving Data to S3 with Apache NiFi

Moving data to the cloud is one of the cornerstones of any cloud migration. Apache NiFi is an open source tool that enables you to easily move and process data using a graphical user interface (GUI).  In this blog post, we will examine a simple way to move data to the cloud using NiFi c...

Read more
  • AWS
  • S3
Avatar
Chandan Patra
— June 11, 2019

Amazon DynamoDB: 10 Things You Should Know

Amazon DynamoDB is a managed NoSQL service with strong consistency and predictable performance that shields users from the complexities of manual setup. Whether or not you've actually used a NoSQL data store yourself, it's probably a good idea to make sure you fully understand the key ...

Read more
  • AWS
  • DynamoDB
Avatar
Andrew Larkin
— June 6, 2019

The 11 AWS Certifications: Which is Right for You and Your Team?

As companies increasingly shift workloads to the public cloud, cloud computing has moved from a nice-to-have to a core competency in the enterprise. This shift requires a new set of skills to design, deploy, and manage applications in cloud computing. As the market leader and most ma...

Read more
  • AWS
  • AWS Certifications
Sam Ghardashem
Sam Ghardashem
— May 15, 2019

Aviatrix Integration of a NextGen Firewall in AWS Transit Gateway

Learn how Aviatrix’s intelligent orchestration and control eliminates unwanted tradeoffs encountered when deploying Palo Alto Networks VM-Series Firewalls with AWS Transit Gateway. Deploying any next generation firewall in a public cloud environment is challenging, not because of the f...

Read more
  • AWS