Responsiveness Under Working Conditions
summary
Defines Responsiveness under working conditions as HTTP/2/3 latency measured while saturating the link with concurrent HTTP bulk transfers, sending HTTP GET of a one‑byte object both on separate and on-load connections to capture multitasking and agility. Idle ping shows ~6 ms but loaded latency can reach ~3 seconds; fixing network buffers with L4S ECN brings latency to single‑digit ms while keeping throughput, and reducing server transmit/TLS/HTTP2 buffering (TCP_NOTSENT_LOWAT, app-layer buffers) cuts video stalls ~20%. Tools: macOS Network Quality, Go responsiveness repo, iperf2, Ookla.
AI-written summary · about
chapters (15)
- 0:00Responsiveness in networks
- 1:12Good internet performance
- 2:57Traditional latency problems
- 4:23Responsiveness under working conditions
- 6:53Measuring responsiveness realistically
- 8:57Tools for network responsiveness
- 10:46Latency spikes under load
- 12:52Network bottlenecks
- 15:20L4S improves latency and throughput
- 17:06Improving network responsiveness
- 18:11Server-side responsiveness discovery
- 19:58Server buffers impact
- 21:43Benefits of server responsiveness
- 22:46Optimizing server responsiveness
- 23:06Responsiveness and user experience
related talks
45mReducing Network Latency: Innovations for a Faster Internet
16m"Latency: Why You Should Worry and What You Can Do About It"
52mPanel 2: Network
26mTuning Your Cloud: Improve Global Network Performance for Applications
23mThe Short and Happy Lives of TCP and HTTP Requests
17mHow to diagnose nanosecond network latencies in rich end-host stacks
41mPerformance Story: How to Speed Up HTTP/2 Client
26mTuning Your Cloud: Improve Global Network Performance for Application