Bringing Decentralized Search to Decentralized Services
Talkat 15th USENIX Symposium on Operating Systems Design and Implementation (OSDI '21) · 2021-07-14..07-16·
summary
DeSearch builds a decentralized, verifiable, private search pipeline for Steemit and OpenBazaar, using trusted hardware (SGX) and ORAM for privacy, cloud storage plus blockchain for integrity, and a network of stateless desktop workers. It stores 80 million records (240 GB), processes 128 million queries/day, and serves single-keyword searches with ~50 ms latency. Verification uses a witness tree per pipeline committed to an on-chain-hashed Kanban so dedicated verifiers drop client verification from 4.5 hours to ~1.2 seconds.
AI-written summary · about
chapters (8)
- 0:00Problem statement
- 2:16Search pipeline challenges
- 4:10Decoupling state and compute
- 5:17Verifiable kanban storage
- 6:01Verifiable lambdas
- 9:05Privacy protections
- 10:15Performance evaluation
- 12:44Conclusion
related talks
22mSharding 2.0: Unlocking Infinite Search Scalability in OpenSearch
17mSPEEDEX: A Scalable, Parallelizable, and Economically Efficient Decentralized EXchange...
36mHow to write a search engine in 15 lines of code
42mIntroduction to DEKU, a Parametric Sidechain on Tezos
31mSharding 2.0: Unlocking Infinite Search Scalability in OpenSearch
42mBeyond Blockchain: Convergent Consensus
31mUber’s Innovations in OpenSearch for a Cloud Native, Serverles...
44mScaling OpenSearch at Uber: Operating Resilient, Fault-Tolerant, High-Performance Clus...