Toban Zolman
August 13, 2015

As you’re probably already aware, one of the best ways to cut down your AWS spending is through AWS Reserved Instances. But, without a solid understanding of RI fundamentals, you can end up spending a lot of money without much return. To start maximizing your AWS cost savings, let’s start with the basics. What is an RI? Why use them in the first place? And what’s all this talk of RI types?

What is an RI?

Amazon Reserved Instances (RIs) allow you to reserve Amazon resources and capacity for 1 or 3 years, in exchange for a significantly discounted hourly rate (up to 75%), compared to On-Demand pricing. RIs are available for a range of Amazon services including EC2, DynamoDB, Redshift and others.

Why make a reservation?

There are several primary reasons to use RIs.

First, RIs are an excellent tool for cost savings; using them can lower the cost of resources you’re already using, by allowing you to pay a lower effective rate than the price you would pay on demand. There are more than 2,000 types of RIs, each with their own break even point— the point where you have gotten enough usage out of the reserved capacity lower rate to make up for any upfront cost you paid. Generally speaking, your savings can be significant by using an RI.

Second, using an RI can lock in future capacity in a region or availability zone. This is useful if you experience spikes in usage. When you purchase an RI, you’re ensuring a lower price over the duration of the reservation— it’s the same hourly cost no matter what. But you’re also have priority access (albeit not guaranteed) to an instance type in that availability zone at any time during your reservation. If you have volatile usage, this is a really useful feature, because you know that Amazon will have capacity for you.

Third, RIs allow you to reserve capacity in another region. Perhaps your service requires regularly scheduled failovers between regions without incurring any sort of outage. If you have a reservation or reservations in that destination region, you will have priority access to capacity (again, not guaranteed) which will help mitigate impact to your service.

What are RI types?

Depending on the AWS service, there are different types of reservations (and corresponding cost savings). For EC2 three types of reservations are offered: “All-Upfronts”, “Partial-Upfronts” and “No-Upfronts”; the first 2 have an upfront fee and the last 2 types have an hourly usage fee. You break even based on their effective utilization. It’s important to remember that with all RI types, each month you are committed to be billed an entire month’s worth of hours. For example, in the case of EC2 Partial-Upfront RIs, this means that you’ll notice a large spike in costs at the beginning of each month; you’re being billed in advance for 744 hours. As you go through the month, you make up for those charges in usage. If you use enough, then you can save a lot of money. But if you don’t use enough, you may actually be spending more than you would for on demand pricing.