Case Study: Roku achieves 32% vCPU reduction across 110+ Kubernetes clusters with ScaleOps

A ScaleOps Case Study

Preview of the Roku Case Study

Roku cuts vCPU waste by 32% across 110+ Kubernetes clusters with ScaleOps

Roku, a major streaming platform, faced significant Kubernetes resource waste across its 110+ clusters. Their engineering teams, wary of causing production incidents, consistently over-provisioned resources, leading to high cloud costs. They turned to the vendor ScaleOps and its automated resource optimization platform to address this challenge without disrupting engineering workflows.

ScaleOps implemented a safety-first solution that automated the rightsizing of CPU and memory allocation. The rollout began with a read-only observation period to build trust before enabling automation. The results were a 32% reduction in vCPU usage at only 20% cluster coverage, with a projected savings of up to 72%. Importantly, ScaleOps achieved this with zero production incidents and no changes required from Roku's engineering teams.


View this case study…

ScaleOps

14 Case Studies