Cloud Machine Manager Blog

Using Reserved Instances to Save Money

Reserved InstancesFor many organizations, be it a start-up or a large corporation, a key strategic goal is often saving money. But finding areas within the organization where savings can be made can be very tricky.

The solution might not necessarily be to reduce budgets, but to find ways to be more efficient with the budget you currently have. For example, when we look at the IT department, adopting a cloud infrastructure is often the first step to increasing efficiency in both costs and departmental processes i.e. reduced spending on infrastructure and improved access to information across users.

And there are ways that the IT department can continue to stand out as an efficient member of the organisation. One option for AWS EC2 users is to consider using Reserved Instances (RI). Essentially, with Reserved Instances you can reserve EC2 computing capacity for 1 or 3 years, enjoying huge cost savings.

Reserved Instances pricing models

Reserved Instances come in 3 different pricing models; All upfront, where you pay for the entire instance fee (1 or 3 years), Partial Upfront, where you pay a partial amount of the fee for the reserved instances, supplemented by monthly payments to cover the remaining fee, and No Upfront, where you pay for your instances in monthly instalments.

So in addition to enjoying the benefits of launching instances whenever you experience spikes in activity, you can also achieve cost-savings of up to 75% compared to On-Demand Instance pricing. So by paying upfront or in monthly instalments, you can gain a significant reduction in the hourly cost of running your servers.

Cost savings of Reserved Instances

For example, for standard Reserved Instances, if you ran an m4.4xlarge server in US East (N. Virginia) on a Linux server, the On-Demand price would be $0.958 per hour. But if you were to pay $4828 All Upfront for a 1-year term, your effective hourly price would be $0.551, offering a 42% saving over On-Demand.

And your savings over On-Demand increase significantly if you were to purchase your Reserved Instances on a 3-year term. By paying $9387 All Upfront, the effective hourly rate drops significantly to $0.357, providing a 65% saving over On-Demand.

If you decided to go for the No Upfront payment option, you could pay $480.34 per month over the course of 1 year. This equates to an effective hourly cost of $0.658 per hour, providing a 31% saving over On-Demand.

This is just one example of how Reserved Instances can help you to save money.

Cost savings by location

Although Reserved Instances are a great way to reduce your costs, you need to be aware of the specificity of cost savings by AWS region.

If we take the previous example of an m4.4xlarge server in US East (N. Virginia) on a Linux server but place this server in US West (Northern California), the On-Demand hourly cost goes up to $1.117.

The All Upfront cost for a 1-year term then increases to $6655, with the effective hourly rate going up to $0.759. Although you’re still making a saving, the saving over On-Demand now drops to 32%.

Equally, if you chose the No Upfront cost path, the effective hourly rate is $0.907 with a saving of 19% over On-Demand – a lot lower than the 31% your competitors over in Northern Virginia are saving.

Don’t get me wrong. From these examples, it’s clear to see that Reserved Instances are a great way to make cost savings across the globe. But it’s important to consider the idea that the location of your servers can significantly impact the amount of savings you can make.

Ultimately, Reserved Instances are a great way for organizations of all sizes to reduce their cloud computing costs and increase their efficiency. So rather than having to face cuts to your budget, it’s worth considering how you can integrate the use of Reserved Instances into your cloud computing infrastructure.