Blazing-Fast Container Deployments - Image Caching With Block Storage
CNCF [Cloud Native Computing Foundation] via YouTube
Learn EDR Internals: Research & Development From The Masters
Learn the Skills Netflix, Meta, and Capital One Actually Hire For
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
Learn how to dramatically accelerate Kubernetes container deployments through innovative image caching techniques using remote block storage in this 25-minute conference talk from CNCF. Discover how traditional container image pulls from remote registries create deployment bottlenecks, particularly for large AI/ML model images, and explore a groundbreaking solution that leverages block storage as a backing store to cache container images across all cluster nodes. Examine the complete architecture and implementation steps for this caching system, while reviewing real-world performance data from NVIDIA Cloud Functions production deployment that achieved over 52% reduction in deployment times. Gain practical strategies for optimizing AI-driven microservices and latency-sensitive applications through cutting-edge storage solutions that enhance both speed and cluster efficiency in fast-paced Kubernetes environments.
Syllabus
Blazing-Fast Container Deployments: Image Caching With Block Storage - Manu Bhadoria & Juana Nakfour
Taught by
CNCF [Cloud Native Computing Foundation]