Skip to main content
Rightsizing

Learn how to find the perfect fit for your cloud resources

Updated over a week ago

Selecting the ideal machine size for your cloud infrastructure can be quite challenging. Often, organizations are uncertain about their exact capacity requirements, leading them to purchase more cloud services than necessary. This impulsive decision-making can result in overcapacity, where the size and type of cloud instances exceed the actual workload demands, resulting in higher cloud bills and budget strain.

The experienced DevOps engineers at Profisea Labs have devised a solution to implement the rightsizing process without adversely affecting your business's financial health. Uniskai scrutinizes the current resource utilization and proposes optimizations to align instance types and sizes with your actual workload capacity.

When you click on 'Rightsizing,' you'll find a list of the instances we work with, such as EC2, ECS, Volume, Lambda, and EKS, along with rightsizing recommendations based on your specific usage. Additionally, you can access detailed information, including instance names, tags, region, type, and usage details.

For instance, in the image below, there's a recommendation to downsize an instance from t3.medium to t3.small, primarily reducing memory allocation. This indicates that the memory isn't fully utilized and can be scaled down while keeping other attributes the same. Even by implementing this single change, you can already enjoy a 30% cost reduction.

If you click the gray arrow on the left, the suggestion will expand, offering alternative rightsizing options for your resource. This allows you to choose the best-fitting recommendation for your needs.

On the right, you'll also find an estimate of the potential cost savings achievable if you accept all the recommendations.

If you decide to implement a recommendation, simply click 'Modify.' If you prefer to postpone a decision, you can remove it from the recommendation list by clicking 'Archive.'

On the 'Rightsizing' tab, you can view and apply recommendations for volumes with the gp2 type. To add a volume recommendation to the 'RS' tab, the instance must be in a 'Running' state for at least 12 hours, and the volume must be of type gp2.

Rightsizing also allows you to adjust Lambda concurrency with a single click. You can view general information during the modification of Lambda functions and make changes to multiple Lambda assets simultaneously.

Uniskai investigates the current usage of particular EBS resources and suggests how they can be optimized to match the types and sizes of the instances to the actual workload capacity usage, with no harm to your business efficiency.      

Our customers can also modify EKS resources. Choose the most suitable option and click the 'Edit' button. Confirm the action, and you'll see the results. To learn more about EKS, click here.

Click on the three dots to view the resource on one of the interactive diagrams in Cloudview.

After clicking 'Modify,' you'll see the 'Modify Resource' dialog below, which displays how the resource details will change after the modification. To accept the suggestion, click 'Modify.'

You can also make bulk modifications by selecting the required resources and clicking 'Modify All.'

Additionally, this table offers filtering capabilities. Click 'Filter,' select the region, VPC item, and action item, and add a tag if necessary. Once you're done, click 'Apply.'

Efficiently manage resources in several accounts at once. Head over to the ‘Rightsizing’ tab, choose the necessary accounts and resources, and make the necessary adjustments. You can take advantage of all the tools designed for a single account, including sorting, searching, filtering, and bulk actions. Save money quickly.

Rightsizing is a pivotal of Uniskai’s strategy for optimizing cloud spending. Our approach identifies all underutilized resources and provides rightsizing recommendations to achieve cost savings in the cloud. We suggest instance sizes and types that align with your workload capacity, potentially resulting in savings of up to 75%.

Did this answer your question?