Architecture of VKontakte's storage
summary
Description of VKontakte’s storage architecture for trillions of photos, audio and other files across thousands of servers. Covers the evolution of their storage design over 18 years, current technologies and approaches used, and how they provide fault tolerance for such large-scale data.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 7:48Legacy file storage
- 10:16Modern file ID approach
- 14:10Two-level storage
- 19:56Erasure coding
- 28:03Large files handling
- 31:15Summary
- 31:58Q&A
related talks
54mEverything Everywhere at Once: Tracing under High Load with 10k Backends
53mFrom Algorithm to Production: How to Approach Verification of Distributed Systems
50mAvito Messenger Architecture – The Journey of a Single Message
12mFacebook's Tectonic Filesystem: Efficiency from Exascale
57mDistributed Systems at Odnoklassniki
53mCheaper, More Reliable, Simpler
46mScaling YouTube's Backend: The Vitess Trade-offs
01h 01mPanel: Facebook's Distributed Database Evolution