Exclusive Whitepaper
Navigating the FinOps Landscape: A Comprehensive Market Analysis

Unconventional Strategies for Optimizing AWS Costs in Non-production Workloads

Shivam Uniyal
By Shivam Uniyal
25 Aug, 2023

AWS Cloud has provided a scalable, cheaper, and easier way of managing the infrastructure with the help of which corporations can easily set up their IT infrastructure in the cloud without very much less hassle and zero upfront cost. 

It provides numerous AWS cost optimization services with the help of which you can easily manage your day-to-day IT tasks, But these services come with additional costs that can increase your bills. But with the right strategy, AWS cloud cost optimization can help your business grow. There are different aspects of the AWS cloud by which you can save your money without compromising on efficiency.

Some of the AWS cost optimization best practices

Scheduling on/off Time for Resources

For non-production environments, you can use automation tools that can switch off unused EC2 Instances and other services that are not in use. For example, non-production environments are often not used at night. Hence, if not switched off, they will generate costs for the whole period. 

Some of the resources that can be scheduled to start and stop are:-

  • Ec2 Instances
  • Relational Databases(RDS)
  • Elastic Container Services
  • Elastic Kubernetes Service(EKS)
  • DocumentDb
  • Auto Scaling Groups(ASG)

How can you schedule on/off time on these resources?

There are many automation-driven AWS cost optimization tools and services available that can be used for scheduling.

Let us take a look at some of the examples.

Jenkins
Jenkins is a powerful CI/CD Tool that can be used to automate the scheduling process with the help of cron. You can set start and stop times for the days for which you want to perform the activity.

Lambda
Lambda is an AWS service that can also be used to schedule start-stop time, but lambda comes with an additional cost.

Cron
You can use cron to run scheduling scripts for desired timings. It is one of the easier ways to create a scheduling strategy.

Using start-stop scheduling of AWS resources will help in AWS cost optimization as there will be a time when these resources will not be used, but they will be billed for their running time.

Using the Right Instance Family

AWS provides Ec2 Instances based on different architectures like AMD and ARM, so we have to select the instances based on our needs and cost strategy.

These instances come with different costs. For example, arm instances are cheaper than AMD instances. Moving your Non-production architecture to ARM-based will help you save AWS costs without compromising performance.

Always choose the instance family after thoroughly comparing the pros and cons and try to choose the latest generations.

Using spot instances for ECS/EKS services auto-scaling groups

For a non-production environment, you can use spot instances, which can help you save huge amounts of money.

What are spot instances?

Amazon sells its server Spot. This means Amazon sells its leftover server space that it has been unable to sell without using a data center. The server is the same server that they provide with the on-demand option. The significant difference is that Amazon can request the server back at 2 minutes' notice (this can cause your services to have an interruption). On the other hand, Amazon pricing optimization can help reach a discount of up to 90%. In most cases, the chances of them asking for the servers back is very low (around 5%).

You can use spot instances by creating a spot fleet in your Auto-scaling groups and selecting the spot fleet instances based on your service requirements.

Use the minimum instance type required for the service type

Always select the instance type based on the service requirement, check the amount of memory and CPU that the service will use, and then select the instance type.

If you do not use the right type of instance for your service, it may be possible that the instance is not being utilized to its fullest capacity, and you will be adhering to the cost of the instances.

Use 50% volume size as compared to the production environment

While using the volumes for EC2 instances always uses 50% less capacity than what is being used for the production environment. Because production environments handle more amount of data than non-production, it will be a wastage of resources if we use the same volume as production.

Don't use clusters in a non-prod environment

Some services, like Hazelcast, Rabbitmq, etc., need clusters when the amount of data is high. But in the case of a non-production environment, we can use a single node only because the non-prod environment is used for testing only, and clustering the resources would generate additional costs for the infrastructure.

Replica of application to one until it is required by the development team for high load

While using EKS/ECS for microservices architecture in non-production environments, always keep the count of replicas to one. For non-prod, a single replica of the service is enough. Using a high replica count will result in additional costs. You can increase the replica count when it is needed by the development team for testing purposes.

Setting backup of Logs for Cloudwatch, S3, etc.

Logs amount to high costs if they don't have a suitable retention period. Storing logs for a large amount of time generates lots of costs.

To bring your bill down, set the retention period of the logs to the minimum desired time that doesn't affect your needs.

Logs keep getting stored and make a huge stack for which AWS charges you a lot, whether they are S3 logs or Cloudwatch logs.

For services like Elasticsearch, storing large amounts of logs affects the performance of the service. It fills the storage of the instance, which can result in system failure or slow response.

Creating desired ECR Image Lifecycle Policy

With a lifecycle policy, you can set rules to automatically delete old or unused images, which can help reduce storage costs. By deleting images that are no longer needed, you can free up space in your registry and avoid paying for storage that is not being used.

Move infrequently-accessed data to lower cost tiers

Moving infrequently-accessed data to lower cost tiers in Amazon S3 can help in s3 cost optimization in several ways:

Lower Storage Costs: Amazon S3 offers several lower-cost storage tiers, such as S3 Standard-Infrequent Access (S3 Standard-IA) and S3 One Zone-Infrequent Access (S3 One Zone-IA), which offer lower storage costs than the S3 Standard storage class. By moving infrequently-accessed data to these lower-cost tiers, you can save on storage costs while still maintaining durability and availability.

Lifecycle Policies: Amazon S3 allows you to automate the process of moving objects to lower-cost storage tiers using lifecycle policies. You can set policies based on access patterns, object age, or other criteria to automatically move objects to lower-cost tiers when they are no longer frequently accessed.

Retrieval Costs: Lower cost tiers in Amazon S3 may have retrieval costs associated with accessing data, such as S3 Standard-IA and S3 One Zone-IA. However, if you are not frequently accessing the data, the savings from the lower storage costs can offset the retrieval costs.

Durability and Availability: Amazon S3's lower cost tiers offer the same level of durability and availability as the S3 Standard storage class, ensuring that your data remains secure and accessible.

Improved Data Management: Moving infrequently-accessed data to lower cost tiers in Amazon S3 can help you better manage your data by allowing you to tier your data based on access frequency and optimize costs based on the value of the data.

Moving infrequently-accessed data to lower-cost tiers in Amazon S3 can help you optimize your storage costs by reducing storage costs, automating the process of moving objects to lower-cost tiers, maintaining durability and availability, and improving data management. However, it is important to carefully evaluate your storage requirements and access patterns to determine which storage tiers are appropriate for your data.

These Methods can help you to bring your AWS infrastructure cost down by a high margin.

Optimize your AWS Costs and achieve enhanced cloud cost savings with CloudKeeper by your side. With contractually guaranteed savings, free access to AWS cost analytics platform, and recommendations from certified AWS experts, CloudKeeper can help reduce your overall AWS bills by up to 25%. Talk to our experts today. 

 

0 Comment
Leave a Comment