Flink on Karmada: Building Resilient Data Pipelines on Multi-Cluster K8s
CNCF [Cloud Native Computing Foundation] via YouTube
The Most Addictive Python and SQL Courses
AI Engineer - Learn how to integrate AI into software applications
Overview
Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
This conference talk explores how to build resilient data pipelines using Apache Flink on Karmada across multiple Kubernetes clusters. Learn how Bloomberg's Streaming Analytics team collaborated with the Karmada community to enhance failover capabilities for stateful workloads. Discover the challenges of intelligently scheduling stateful applications while maintaining state consistency during failovers in multi-cluster Kubernetes environments. The presentation covers open source enhancements to Karmada that address these challenges and demonstrates how to apply these solutions to other stateful use cases beyond Apache Flink. Gain practical insights into improving workload resiliency while preserving critical state information for data processing applications.
Syllabus
Flink on Karmada: Building Resilient Data Pipelines on Multi-Cluster K8s - Michas Szacillo & Wang Li
Taught by
CNCF [Cloud Native Computing Foundation]