7 Ways to Save on Your AWS Bill

Save AWS Cost

Research by Gartner concluded that companies spend two to three times higher than forecasted for their public cloud bills every year. In reality, this was a tall claim back then. Now, as we enter into the post-pandemic era where cloud investments are gold, there are chances that the statement may be true. There's little doubt that in the face of uncertainty, millions of dollars were and are being spent on unnecessary cloud costs as businesses fail to take steps to minimize AWS bill costs.  

This may not entirely be in the hands of the business itself. AWS pricing is complicated; and, if a business believes that they only pay for what they use, cloud bills can accumulate and exceed budgets. Added to this, are extra services attached to instances. These drive costs up—and can continue to do so even when instances end.   

AWS has a wide breadth of services and pricing options that offer the flexibility to effectively manage your costs while maintaining performance. While the fundamental process of optimizing AWS costs is obvious - monitor usage, analyze data to save, and implement to realize savings; in this blog, we explore a more tactical approach to reducing cost and saving on your AWS bill.  

1.  Regularly Delete EBS Volumes

When an EC2 instance is launched, an Elastic Block Storage (EBS) volume gets attached.  Unless the box in the AWS console automatically deletes the volume when the instance is terminated is checked, AWS will continue to charge for it even though it isn't being used. Find the unattached EBS volumes that you no longer need, and then delete them after optionally creating an EBS snapshot. Important: The data in the EBS volume is lost on deletion, and the volume can't be attached to any instance. 

Use Amazon EC2 Spot instances for your stateless, fault-tolerant, or flexible applications, such as containerized workloads, to reduce your operating costs. 

2. Take A Snapshot of the Inventory 

On their own, each EBS snapshot is inexpensive. However, when obsolete snapshots accumulate, unnecessary costs pile up too. Most often only the most recent snapshot is needed to aid recovery. Hence, maintain a threshold for retaining snapshots and delete the rest to save on your AWS bill.  

3.  Monitor Elastic IP addresses  

The pricing of Elastic IP addresses has been structured in favor of AWS. These are free as long as they are attached when they run. The moment an instance is stopped, the unused IP starts piling up costs. The costs add up quickly, and since Elastic IP addresses are difficult to find in AWS System Manager or AWS Console, they go unnoticed. AWS Trusted Advisor can help you identify resources that are not being utilized to their full capacity. Then, you can decide to delete these underutilized resources to optimize your costs. 

4.   Schedule Instances Accurately 

There is a misconception that one pays only for what one uses in AWS. When it comes to assets that are used for development, testing, and staging, it is important to note that they should not be left running when they aren't being used. Depending on the aggressiveness of schedules, there is potential to save more than 65% by closely monitoring the start and stop of non-production assets.  

5.  Buy Reservations for Redshift and ElastiCache Services 

Redshift and ElastiCache are two of the services for which you can purchase reservations to help reduce your AWS bills. Reserved nodes can be purchased for 1-year or 3-year terms, with the option of paying the full amount upfront or partially upfront or paying monthly. A reminder though- in order to utilize the ElastiCache Service best, one must first upgrade Nodes to the latest generations. If you have predictable workloads on Amazon Elastic Compute Cloud (Amazon EC2), AWS Fargate, or AWS Lambda, don't pay the default On-Demand pricing, you can save money by choosing an appropriate Savings Plan. 

6.  Set Retention Periods on CloudWatch Logs  

CloudWatch logs can cost more if a high amount of data is constantly getting pushed to the log stream. You can save the storage cost by configuring a predetermined retention period for logs, a basis on which they will get automatically deleted.  Retention periods can be configured from 1 day to 2 years. 

7.  Set Lifecycle Rules for S3 and ECR 

S3: 

S3 is likely to hold objects which are not frequently accessed. These can be configured to move to a different storage class within S3, which bills very low for storage. S3 Standard-IA, S3 One Zone-IA, and S3 Glacier Instant Retrieval are some storage classes that can save you money. You need to choose one based on your needs and access patterns. 

ECR: 

The storage usage in the repository is directly proportional to your billing. Hence, when more docker images with different tags are stored, your bill increases. This can be avoided by ensuring that a lifecycle period is configured for docker images such that they get deleted automatically according to timestamp or when new images are used.  

Optimizing and saving AWS bill costs shouldn't be just a periodic exercise. For the best possible results, a business's cloud environment will have to be monitored constantly to identify unattached, unused, and underused assets and manage them accordingly. This is where KnackForge steps in.   

Wondering how we can help you manage your cloud, save on AWS and scale your business? Simply fill out this form or get in touch with us at sales@knackforge.com