On Data Locality in Kubernetes
Explore the challenges and solutions surrounding data locality in Kubernetes in this informative conference talk. Delve into the complexities of accessing data from cloud-native sources in Kubernetes environments and understand why platform engineers often resort to copying data for I/O throughput optimization. Examine various approaches to emulate or introduce data locality in Kubernetes schedulers, weighing their advantages and disadvantages. Gain insights into the importance of addressing data locality for improving efficiency in data-intensive workloads within Kubernetes ecosystems. Learn from experts Chen Wang from IBM Research and Shouwei Chen from Alluxio as they discuss the future implications of data locality for Kubernetes performance.