Completed
Impact the performance of elastic applications
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
KRCORE - A Microsecond-scale RDMA Control Plane for Elastic Computing
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Remote Direct Memory Access (RDMA)
- 3 Problem: creating RDMA connections is slow
- 4 No problem for traditional applications
- 5 Impact the performance of elastic applications
- 6 Basic idea: connection pooling & reusing
- 7 User-space QPs cannot be shared
- 8 Solution #1. share QPs in a kernel-space QP poc
- 9 Massive QPs cached in the pool
- 10 Solution #2. Retrofit DCT as the shared connecti
- 11 Our design: MetaServer
- 12 Correct QP multiplexing
- 13 We add additional checks to prevent QP corrupt
- 14 Put it all together: KRCore
- 15 KRCore implementation
- 16 Evaluations setup
- 17 Control plane performance of KRCore
- 18 Data plane performance
- 19 Accelerating disaggregated RDMA-enabled KVS
- 20 Summary and discussion
- 21 Conclusion of KRCore
