Testing Performance of Distributed Storage Systems
summary
Covers performance testing of Software Defined Storage (SDS) systems by examining SDS architectures, cache layers, the high cost of sync operations, common bottlenecks, and the factors that actually affect throughput and latency. Describes a detailed testing methodology and also discusses data consistency and reliability trade-offs, noting that some SDS implementations sacrifice durability for speed.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:58Parallels experience
- 4:31Testing goals
- 7:46Tools and workloads
- 10:00Test statistics and conditions
- 11:27fsync importance
- 13:35fsync scalability
- 15:04fsync testing challenges
- 17:15Parallels Cluster methodology
- 20:39Data consistency testing
- 24:42Caching and hardware
- 26:53SSD reliability
- 27:37MTBF
- 29:53Recovery after failures
- 31:22Q&A
- 41:11Comparison with others
related talks
50mPerformance comparison of distributed file systems
38mLoad Testing Storage Systems and the Peculiarities of Generating Test Data
47mA Distributed, Scalable and Highly Reliable Data Storage System
21mNew Techniques to Improve Small I/O Workloads in Distributed File Systems
40mBenchmarking: Repeatable & Comparable
25mDistCache: Provable Load Balancing for Large-Scale Storage Systems with Distributed...
40mPerformance Testing and Big Data
34mDistributed systems 102