Introducing Spot Wave: Serverless Spark on Kubernetes

Wave delivers automated, reliable, cost-efficient, and continuously optimized cloud infrastructure specifically for big data Spark applications:

• Optimization engine that automatically chooses the best infrastructure to run a Spark job
• Run Spark applications reliably on a mix of spot, on-demand and reserved instances
• Spark application rightsizing that tunes configurations based on actual resource consumption to limit over-provisioning
• Container bin-packing that ensures pods are allocated to nodes efficiently, reducing underutilization of compute
• Pre-built integrations with big data tooling to meet data engineers and scientists where they’re at