~/talk

Hyperscan: A Fast Multi-pattern Regex Matcher for Modern CPUs

Talkat NSDI 2019 · 2019-02-26..02-28· [networking]
links:none on file
name:Yang Hong
links:none on file
links:none on file
links:none on file
links:none on file
name:Jiayu Hu
links:none on file
links:none on file
slides:none on file

Hyperscan decomposes complex regexes into string fragments and small FA components using graph analyses (dominant path, region, max‑flow/min‑cut), compiles tens of thousands of rules in under a second, and implements graph decomposition plus SIMD‑accelerated shift/bitset DFA engines. It replaces two‑stage manual prefilt ers to avoid duplicate scans, yields about 3x–13x component speedups and an 8.7× Snort throughput improvement on real traces, and is production‑deployed in 40+ commercial products and many cloud servers.

AI-written summary · about

  1. 0:00Intro
  2. 0:24Networking Applications with Regex Matching
  3. 3:16Contributions
  4. 4:05Wide Adoption of Hyperscan
  5. 4:43Decomposition-based Matching
  6. 6:40Key Issues with Regex Decomposition
  7. 7:06Graph-based Regex Decomposition
  8. 9:14Effectiveness of Graph Analysis on Real-world Rules
  9. 9:44Quality of Automatically Extracted Keywords
  10. 10:43How to Accelerate Pattern Matching Algorithms?
  11. 11:15Multi-string Pattern Matching Overview
  12. 12:51Multi-string Shift-or Matching
  13. 15:42Other Subsystems
  14. 16:12Evaluation of Hyperscan
  15. 17:19Repex Matching Performance
  16. 17:45Real-world DPI Application - Snort
  17. 18:33Conclusion
~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.