The power and flexibility of Kubernetes for Spark applications—without the complexities

Applications run on the lowest-cost, highest-performing infrastructure possible.
Data teams can focus on building data applications, not managing cloud infrastructure or debugging Apache Spark issues.
Benefit from an open and flexible containerized architecture, leveraging the best open-source technology without the complexity.

Ocean for Apache Spark

Run Spark without managing servers

Big data workloads run reliably on optimized infrastructure that’s been automatically provisioned with an optimal mix of spot, on-demand and reserved instances.

Make Spark-on-Kubernetes developer-friendly

Setup, configuration, maintenance and scaling of Spark applications and Kubernetes clusters are easy with intuitive UIs, key integrations and powerful automations.

Your cloud, your Kubernetes, your tools

Apply your security and data privacy best practices by deploying on K8s in your own cloud account and VPC. Leverage your data tools by integrating your Jupyter notebooks, IDEs, and schedulers.

Kubernetes toolset
Big Data
The United Nations modernized their martime traffic data exploration while cutting costs by 70%.
Read Case Study

Key features

Spark-centric monitoring and observability
Track your application’s configurations, logs, Spark UI and key metrics in our live and historical dashboard. Leverage Delight, our unique Spark-centric observability layer, to easily identify issues and improve performance of your workloads.
History-based Spark configuration tuning
The container sizes, number of executors and Spark configuration flags of your recurring pipelines are automatically tuned based on their historical performance to optimize their performance and reliability and ensure that they finish within their SLA.
Spark-aware infrastructure scaling
Kubernetes clusters are automatically scaled while maintaining dynamic headroom to ensure applications can run instantaneously. Instance types and container sizes are determined to satisfy your workload requirements while maximizing bin-packing.
Run reliably on spot instances
Save on cloud compute by leveraging spot instances with Ocean’s enterprise-grade SLA for performance and reliability. Track your cloud costs, broken down by each user and each job, in real-time within our dashboard.
Leveraging Ocean for Apache Spark decreases cost while letting us sleep well at night and achieve the plans we dream about.
Dale McCrory, Co-Founder
Lingk reduced their total cost of ownership by 65%!
Learn How


Our serverless pricing only kicks in when Spark cores are running. There’s no rounding, no hidden fee, and the support from our team of Spark experts is included.


Get all of Ocean for Apache Spark features and optimizations

$0.025 per
Spark core hour

Month-to-month with
no annual commitment

Schedule a Demo

Enterprise plan

Custom integrations and customer success packages

Price depends
on volume

Yearly contract

Schedule a Demo

Get started

Ocean for Apache Spark is in preview