7 Ways To Reduce Your AWS Bill

Conquer cloud economics and show me the money

Most companies are spending way too much on compute and storage than necessary — especially on capital expenses for excess capacity in their on-premises data centers required to support peak demand.

The shift to public cloud providers such as AWS allows companies to increase the utilization rates of their workloads, while simultaneously reducing their overall costs. Substituting traditional up-front hardware purchases with the more efficient pay-as-you-go model offers significant advantages.

The massive scale at which AWS operates also benefits their customers — ever increasing economies of scale enables them to continually reduce the pricing of compute and storage services. For example, AWS has reduced the per GB storage price of S3 by 80% since the service was first introduced in 2006.

AWS can drastically change your economic model of running infrastructure and platform services — but only when used correctly.

When building applications and workloads on AWS, you need to take control over the economic model of your architecture. It’s important to think beyond the basic pricing advantages compared to on-premises data centers, and investigate how to effectively use the services to lower your AWS bill.

There are seven ways to optimize your AWS costs:

#1: Shutdown Unused AWS Resources
To optimize cost, it’s critical to shut down unused instances — especially in development environments at the end of each working day and weekends. Services such as AWS OpWorks and Elastic Beanstalk allow developers to quickly deploy and redeploy applications with consistency, without worrying about the configuration of the underlying infrastructure.

By using AWS CloudFormation to define your infrastructure as code, it’s very easy for developers to create resource templates of your AWS resources to quickly build and rebuild your environments. This approach to cloud computing enables the core concept of ephemeral workloads, making it possible to shutdown and delete unused AWS resources without concern.

#2: Use the Appropriate Storage Class
The are 5 tiers of Amazon S3 object storage available, and it’s important to know why and when to use each class to optimize your costs. For each tier, the cost is broken down into the the actual storage amount, the number of HTTP PUT requests, the number of the HTTP GET requests, and volume of data transferred.

To optimize the cost of you data storage, consider implementing object lifecycle management that automatically transitions data between the storage classes. For instance, you can automatically move your data from S3 Standard to IA after 30 days, archive data to Glacier after 90 days, or setup a delete policy to expire specific objects after 180 days.

#3. Select the Right Instance Type
Since different instance families cost different amounts, it’s important to ensure that you are using the most cost-effective instances . Be sure to select the instance that best suits your application workload.

To maximize your workloads while minimizing your spend, consider your specific use case when determining factors like the type of processing unit and and amount of memory required. Optimize the instance resource that results in the delivery of price performance for the price. At least twice a year, assess your choice of instances to ensure they match the reality of your workload.

#4: Monitor, Track, and Analyze your Services Usage
Trusted Advisor and CloudWatch are monitoring and management tools to access your instance metrics. Based on the data collected, you can assess your workloads and scale your instance size up or down.

Trusted Advisor is an excellent tool since it identifies idle resources by running configuration checks. The services also provides real time guidance to help you provision your resources following AWS best practices — providing weekly updates to increase security and performance and reduce your overall costs.

Amazon Cloudwatch is a helpful service to set alarms, collect & monitor log files, and automatically react to changes in your resources that may contribute to waste. With Cloudwatch, you can also monitor custom metrics generated by your own applications via a simple API by sending and storing metrics that are important to your application’s performance — and then take action to manage inefficient or wasteful workloads.

#5: Use Auto Scaling
One of the key benefits of cloud computing is the ability to align your resources with customer demand. To handle variable demand or sudden traffic spikes, you can design dynamically for capacity by using Auto Scaling — adding additional resources only when required and turning them off when they are no longer needed.

The benefit of applying Auto Scaling to your application’s architecture isn’t just limited to better cost management. The services also enables you to detect when an instance is unhealthy, and then automatically terminate the instance and relaunch a new version.

Setting up Auto Scaling very straightforward:

#6: Consolidated Billing
If you have multiple accounts, getting a single bill is very helpful for managing costs and controlling spend. Consolidated Billing enables you to see a combined view of all your AWS charges incurred across all of your accounts.

The service is available at no additional charge — one of your accounts is designated the Master Account. The Master Account pays the charges that are accumulated by all the other accounts in the consolidated billing family. The charges from each account can be easily tracked, and the cost data can also be downloaded in a CSV format.

A cost saving example:
Let’s consider 2 AWS accounts named Alice and Eve.

Without the benefit of Consolidated Billing, AWS would have charged Alice and Eve individually for their usage, and the volume discount for data transfer would not have been applied — for a total $2726.50. By simple simply using Consolidated Billing, the total cost savings is $166.50!

#7: Use Reserved and Spot Instances
Unlike on-demand instances, a Reserved Instance is a 1 or 3-year commitment to purchasing a reservation of capacity in exchange for a significantly lower hourly rate. By using Reserved Instances, you can save up to 75% over equivalent on-demand capacity. And don’t worry — if you buy a RI and you don’t need it, you can sell it back or buy a shorter duration version in the reserved instance marketplace.

Reservations are available with three different payment options: Full Upfront, Partial Upfront, and No Upfront. With partial and no upfront, you pay the remaining balance monthly over the term. In addition to Reserved Instances, Amazon RDS, Dynamo DB, Redshift and ElastiCache are other services where you can take advantage of reserved capacity at lower cost.

Spot Instances are a another phenomenal way to save money for non-stateful workloads — simply bid on EC2 capacity which is not currently in use. Spot Instances are ideal for workloads where you need access to large amount of compute capacity but you are not concerned about an interruption assuming you have a mechanism for dealing with the interruption. The prices of Spot Instances vary on the basis of current demand.

Once you’ve mastered the basics, there are a lot of additional AWS tools to help manage and control your spend including Cost Explorer, Billing Dashboard, and Detailed Billing Report.

By using these tools and services correctly, AWS can drastically change your economic model and reap the financial benefits of cloud computing.

Related posts:

There is only one source of our salvation. In our culture, we run to all sorts of things to save us — better diets, better educations, better ideologies, better mindsets, better cars. We look to…

In February I proposed The case for Bearer Health Records which in some ways resembles bearer digital instruments and offers the independence of the identity- with the private-key owner always able…

This is something our CEO always says because while everyone is moving quickly to get to market, what often gets lost in the process of product launches and marketing campaigns is a clear…