On QUIC, HTTP/3 and tech stack unification
summary
HTTP/3 moves transport into UDP and combines TLS so connections resume across network switches via connection IDs, removes head‑of‑line blocking so multiple streams are unblocked concurrently, and halves handshake overhead versus TCP+TLS which yields a cited 16% time‑to‑first‑byte improvement (200ms → 168ms) on a 40.3Gb/day deployment. Deployment notes: test with nginx and curl or edge‑terminate HTTP/3 (Apache HTTPD lacks good support), expect higher CPU for some ciphers and reduced operational visibility due to full encryption.
AI-written summary · about
chapters (11)
- 0:00HTTP/3 and QUIC intro
- 1:51HTTP evolution 1.0 to 3
- 8:57Why HTTP needed revision
- 12:04Protocol complexity spiral
- 16:16QUIC for transport unification
- 21:16HTTP/3 performance trade-offs
- 23:46Setting up HTTP/3
- 27:40HTTP/3 real-world impact
- 29:03Q&A: Worth the effort?
- 31:37Q&A: Default timeline?
- 33:45Q&A: QUIC security impact
related talks
51mHTTP/3 and QUIC: Who, what, where, when and, WHY?
46mHTTP/3: It's all about the transport!
37mHTTP/3: why and where are we?
34mHTTP/3 is coming
30mHTTP/3 Is Next Generation HTTP. Is It QUIC Enough?
40mHTTP/3 - HTTP over QUIC is the next generation
34mSecuring the Next Version of HTTP: How QUIC and HTTP/3 Compare to HTTP/2
34mQUIC and HTTP/3: Too Big to Fail?!