Open Source OS Approach Self-Detects Attacks & Self-Restores in Seconds
summary
An open‑source OS prototype puts every OS and app state into a DBMS layer that replaces Linux/Kubernetes, uses change‑capture provenance and spools history to a data warehouse; detection queries run in SQL and flag composite suspicious events in a few hundred milliseconds. The demo shows a compromised admin account being suspended and a ransomwarelike exfiltration rolled back to the preattack state in seconds. Performance matches native systems on VoltDB and PostgreSQL, provenance costs about 5% overhead, and the project is available as an open source prototype.
AI-written summary · about
chapters (9)
- 0:00Introduction and motivation
- 6:53OS as a database system
- 13:28Performance and simplicity benefits
- 19:19Provenance and unified logging
- 27:27Cloud and serverless integration
- 35:31Real-world use cases and security
- 41:33Live demo of attack detection
- 44:48Open source availability and next steps
- 47:39Q&A session
related talks
38mOpen Source - A Blessing or a Curse?
31mFuture-proofing operational data at scale
23mAn Open Source Architecture for Modernizing Market Risk Management
38mWhat we built with Open Source
56mBlended Web and Database Attacks on Real-Time, In-Memory Platforms
19mYou’re Doing It All Wrong
17mOpen Source Database Operations in the Age of AI
26mTransactional Serverless with DBOS