We all know that AWS lets us deploy applications in ways that would have been unimaginable using traditional IT infrastructures and processes. But for that very reason, it’s a bit too easy to lose track of exactly what you have running at any one time. The end of a billing period might consequently bring unexpected AWS costs indicating that your ongoing charges have soared out of control.
To avoid this, you’ll have to better understand where these surprises can come from. In this article, I will describe what I believe to be the five biggest AWS costs control and resources management mistakes people and organizations currently make.
Carefully choosing the correct instance type is a fundamental decision you’ll have to confront. What size instances should you run? How many will you need? How will you keep track of all your resources (a good place to start is your AWS account’s Billing page)? Each infrastructure deployment has its own unique characteristics, so these are questions that only you and your organization can answer. Instance types that are specifically optimized for compute, memory, or storage are obviously going to cost more than low-tier or generic versions. It’s your responsibility to ensure you’re not over-provisioning for your specific project. But remember: using oversized instances will cost you money, using too many instances will cost you money, and leaving unused instances running will, you guessed it, cost you money.
Besides instance types, there are also three distinct ways to purchase instance resources and thereby closely control your AWS costs:
For a first-time AWS user, On-Demand Instances are the easiest to deploy and “no-commitment” sounds attractive. But using On-Demand for any sustained periods can generate unexpected AWS costs at the end of the month.
Continuously stopping and starting these instances will also prove costly. If you are planning on deploying a site that needs to operate 24/7 and expect to run this site over for the next 12 months, you would be far better off with a reserved instance, since you’ll know your exact costs up front. If you’re planning to deploy infrastructure that’s a little more complex and requires Elastic Load Balancers and Auto Scaling, you will probably need to draw on a combination of both On-Demand and Spot Instances, carefully calculating the relative AWS costs. Doing this right requires some careful planning.
You can back up the data on your Amazon EBS volumes to Amazon S3 by taking point-in-time snapshots. Backing up your data is not only a smart thing to do but, more often than not, it’s critical. However, EBS snapshots should be saved in moderation, or you may find yourself suffering from snapshot sprawl…which can exponentially increase your storage costs. Make sure you have an EBS Snapshots retention strategy in place (using S3 Lifecycle rules) that’s a good fit for your particular needs.
AWS allows you one Elastic IP (EIP) address associated with a running instance at no charge. However, you will be charged for each additional EIP that you associate with that instance per hour on a pro rata basis. To ensure efficient use of Elastic IP addresses (which are, by definition, a limited resource), AWS also imposes a small hourly charge when these IP addresses are not associated with a running instance or when they are associated with a stopped instance or an unattached network interface. That’s right: you may just stop your instance and assume that the IP associated with it will be automatically released. Unfortunately, this is not the case, and your accrued AWS costs will prove it. So make sure that you aren’t maintaining (and paying for) any unused EIPs.
Amazon CloudWatch is a monitoring service for AWS cloud resources and the applications you run on AWS. You can use Amazon CloudWatch to collect and track metrics, monitor log files, and set alarms. You can set alarms on any of your metrics to receive notifications or automatically initiate other actions (like shutting down unused or underused EC2 instances) when a metric crosses your specified threshold. You can also use Auto Scaling to add or remove Amazon EC2 instances dynamically based on your Amazon CloudWatch metrics. If you have a complex infrastructure of any kind running on AWS, you really need to use CloudWatch to keep up with it.
Most importantly, however, can be the Billing Alerts you create to warn you when your monthly costs reach a certain predetermined threshold.
Careful consideration of the above points will hopefully help you keep your ongoing AWS costs down. As we mentioned, Amazon is a fabulous platform for deploying large applications at prices unimaginable until recently. However if these applications aren’t carefully monitored and kept in check, you may find that the money you’d hoped to save by deploying your infrastructure on AWS is silently flying out the window.
Read also EC2 Pricing: understanding compute costs on AWS
It's Flash Sale time! Get 50% off your first year with Cloud Academy: all access to AWS, Azure, and Cloud…
In this blog post, we're going to answer some questions you might have about the new AWS Certified Data Engineer…
This is my 3rd and final post of this series ‘Navigating the Vocabulary of Gen AI’. If you would like…