# Fargate resource rounding and per-pod overhead driving step-up costs Canonical: https://www.pointfive.co/efficiency-hub/inefficiencies/fargate-resource-rounding-and-per-pod-overhead-driving-step-up-costs Pod resource requests - often inflated by sidecar containers-push total memory or CPU just over a Fargate sizing boundary. By: Tai Nguyen [Cloud Efficiency Hub](https://www.pointfive.co/efficiency-hub) # Fargate Resource Rounding and Per-Pod Overhead Driving Step-Up Costs The short version Pod resource requests - often inflated by sidecar containers-push total memory or CPU just over a Fargate sizing boundary. [Tai Nguyen](https://www.linkedin.com/in/taind772/) Senior Site Reliability Engineer @ Spartan AWS service [AWS EKS](https://www.pointfive.co/efficiency-hub/cloud-services/aws-eks) Category [Compute](https://www.pointfive.co/efficiency-hub/service-category/compute) Reference CER-0280 Type Overprovisioned Resource ## Explanation Why the waste happens and who it affects. Because Fargate adds mandatory system overhead and only supports fixed resource combinations, small incremental increases can force a pod into a much larger billing tier. This results in materially higher cost for marginal additional resource needs, especially in workloads that run continuously or at scale. ## Billing model The pricing dimensions that drive this cost. Fargate charges per vCPU-second and GB-second based on the provisioned pod size. EKS on Fargate adds a fixed 256 MB memory overhead per pod and rounds resource requests up to the nearest supported compute configuration. ## How to detect 3 checks to find it in your estate. - Review pods whose requested resources sit near known Fargate sizing thresholds - Assess whether sidecar containers materially increase total requested CPU or memory - Identify workloads where provisioned capacity significantly exceeds observed workload needs ## How to fix 3 ways to remove the waste. - Right-size pod resource requests with explicit awareness of Fargate overhead and rounding behavior - Tune or consolidate sidecar resource requests to avoid unnecessary tier step-ups - Consider managed node groups for workloads where bin-packing efficiency outweighs Fargate simplicity ## Documentation Vendor references for pricing and configuration. - [https://docs.aws.amazon.com/eks/latest/userguide/fargate-pod-configuration.html docs.aws.amazon.com](https://docs.aws.amazon.com/eks/latest/userguide/fargate-pod-configuration.html) - [https://aws.amazon.com/fargate/pricing/ aws.amazon.com](https://aws.amazon.com/fargate/pricing/) ## Related inefficiencies [Browse the library](https://www.pointfive.co/efficiency-hub) - AWS EKS CER-0298 ### [Stale Completed or Failed Fargate Pods Causing Direct Billing and Capacity Waste](https://www.pointfive.co/efficiency-hub/inefficiencies/stale-completed-or-failed-fargate-pods-causing-direct-billing-and-capacity-waste) Kubernetes Jobs or CronJobs running on EKS Fargate leave completed or failed pod objects in the cluster indefinitely. Although the workload execution has finished, AWS keeps the underlying Fargate microVM running to allow log inspection... Compute - AWS EKS CER-0120 ### [Orphaned and Overprovisioned Resources in EKS Clusters](https://www.pointfive.co/efficiency-hub/inefficiencies/orphaned-and-overprovisioned-resources-in-eks-clusters) In EKS environments, cluster sprawl can occur when workloads are removed but underlying resources remain. Common issues include persistent volumes no longer mounted by pods, services still backed by ELBs despite being unused, and... Compute - AWS EKS CER-0192 ### [Orphaned Kubernetes Resources in EKS](https://www.pointfive.co/efficiency-hub/inefficiencies/orphaned-kubernetes-resources-e7e68) In Kubernetes environments, resources such as ConfigMaps, Secrets, Services, and Persistent Volume Claims (PVCs) are often created dynamically by applications or deployment pipelines. When applications are removed or reconfigured, these... Compute --- Source: the public page above. Product screenshots and illustrative interfaces are examples, not live customer data.