https://www.infoq.com InfoQ Debugging News feed https://www.infoq.com/news/2026/05/pinterest-cpu-zombies-bottleneck/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=Debugging-news <img src="https://res.infoq.com/news/2026/05/pinterest-cpu-zombies-bottleneck/en/headerimage/header-1778308038640.jpeg"/><p>Pinterest identified and resolved CPU starvation issues that affected machine learning training jobs on its Kubernetes-based platform, PinCompute. The engineers traced the problem to an unused Amazon ECS agent, which caused memory cgroup leaks. By disabling the agent, they stabilised performance. This case illustrates the importance of understanding system defaults for effective troubleshooting.</p> <i>By Mark Silvester</i> Performance Observability Performance Tuning Debugging DevOps Architecture & Design Development news Thu, 14 May 2026 10:00:00 GMT https://www.infoq.com/news/2026/05/pinterest-cpu-zombies-bottleneck/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=Debugging-news Mark Silvester 2026-05-14T10:00:00Z /news/2026/05/pinterest-cpu-zombies-bottleneck/en