Ruby for Distributed Storage System
summary
A distributed storage called bigdam-pool exists as both a Java implementation and a Ruby port aiming byte-for-byte API and on-disk compatibility; full JVM-vs-Ruby benchmarks are planned but not finished. Ruby proves fast for writing mocks, HTTP APIs with Sinatra, and integration tests, but serialization (JSON/ MessagePack) and checksum/replication loops drive high CPU. Missing are polished async worker primitives, timer/queue semantics and monitor introspection, which make production parity harder.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 3:23Performance in storage systems
- 7:00Replication strategies
- 10:19BigDum project overview
- 14:45BigDum storage characteristics
- 20:19BigDum architecture design
- 23:11Ruby for mocks and tests
- 26:48Ruby needs in serialization
- 28:37Async and worker libraries
- 33:59Resource management
- 36:31JSON serialization issues
- 37:39Method overloading
- 38:44Ruby for app servers
- 39:47Future of Ruby in storage
- 40:11Conclusion and Q&A
