Explore a 47-minute conference talk from SNIA Storage Developer Conference 2024 that delves into an open-source distributed block storage system providing standard NVMe-oF interface to hosts. Discover how a single virtual block disk can deliver over 2 million IOPS in TCP NVMe-oF mode without specialized hardware offload. Learn the architecture, advantages, and limitations of this high-speed distributed storage solution, along with its practical applications and Kubernetes integration capabilities. Presented by Amazon's Peng Yu, gain comprehensive insights into building efficient distributed storage systems and understand the technical implementation details of the Distributed-NVMe project, which is freely available on GitHub. Master the fundamentals of distributed storage architecture while exploring real-world use cases and integration scenarios with modern container orchestration platforms.
Building High-Speed Distributed Storage with Distributed NVMe
Overview
Syllabus
SNIA SDC 2024 - Distributed-NVMe
Taught by
SNIAVideo
Related Courses
-
Distributed Data Placement with NVMe-oF - Design and Implementation Challenges
-
High-Performance Block Volumes in Virtual Cloud Environment
-
Storage Acceleration via Decoupled Software Defined Storage Architecture
-
Intel IPU and NVMe Initiator with SPDK - Introduction to Hardware Integration and Acceleration Framework
-
Disaggregated NVMe/TCP Storage Using an Infrastructure Processing Unit (IPU)
-
What Can Storage Do for AI? - Optimizing NVMe Storage for Large Language Models
