aws
[Fargate] [request]: Feedback on Reduced Performance for Fargate Tasks with 0.25 vCPU/0.5 GB after May 2025 Platform Upgrade
Community Note
- Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
- Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
- If you are interested in working on this issue or have submitted a pull request, please leave a comment
Tell us about your request What do you want us to build? Following the Fargate platform version upgrade on May 1st, 2025, tasks using 0.25 vCPU and 0.5 GB memory began experiencing degraded performance. This is a request to review the impact of the change and consider improvements or guidance.
Which service(s) is this request for? Fargate
Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard? After the upgrade, services running on smaller task sizes (0.25 vCPU / 0.5 GB memory) became unstable. Tasks frequently failed health checks, became unresponsive, and were replaced more often. This led to slower response times, timeout errors, and direct impact on user experience. The same configurations were previously stable, and no application or task definition changes were made during this time.
The issue was isolated to small task sizes and larger configurations continued to perform reliably. This change has forced a shift to higher task sizes to restore stability, leading to increased operational costs.
Are you currently working around this issue? How are you currently solving this problem? Yes, task definitions were updated to use 0.5 vCPU / 1 GB memory, which resolved the performance issues but increased cost.
Additional context Anything else we should know? The change limits cost-effective scaling and reduces the competitiveness of the service for certain use cases. This feedback is intended to support future planning and improvement of platform behavior at lower task sizes.
Hi @chiragak , if you have not already, do you mind opening a support case referencing some of the Task IDs you deployed recently on .25 vCPU? Can you then share the case ID on here? Thanks.
https://repost.aws/knowledge-center/get-aws-technical-support
