Tools for eBPF DevSecOps
summary
Research project builds an ebpf program registry that auto-annotates source with human and model-generated summaries, infers capabilities (tunneling, parsing, observability) and indexes them in Elasticsearch + Kibana GUI for semantic search. A Python-based extractor producing extracted.c uses CodeQuery and TXL to pull a function’s call graph, dependencies and licenses, outputting a standalone file that compiles and passes the eBPF verifier if the original did. Cites Facebook running ~40 and Netflix ~15 eBPF programs per server.
AI-written summary · about
chapters (11)
- 0:00Start
- 2:06eBPF: extended Berkeley Packet Filter
- 9:00eBPF: What is hard?
- 11:00eBPF: Code reuse is hard
- 13:37eBPF Program Registry
- 14:28Multi-faceted Code Explainability
- 16:00Elastic Based Architecture For Consumption
- 16:35LLM model performance some numbers
- 17:39Reuse Code
- 20:10Challenge: Minimize Code Debt
- 22:29Use case: Debugging of Deployments
related talks
14mWriting Custom eBPF Programs for O11y: What You Need to Know
40mPanel Discussion: eBPF: A New Era in Cloud Infrastructure Tools
32meBPF: The Next Power Tool of SREs
01h 18mTutorial: Getting Started with eBPF
29meBPF on the Rise - Getting Started
43meBPF Superpowers for SRE
16mWhy Everyone Should Care About eBPF
35mAn Introduction to eBPF - Superpowers for Networking, Observability & Security