10 GB/s from Europe to apache.org in...
1 html, 5 css, 5 js, 6 img, no server push
HTTP with ~0 Latency
2k resource, 500k x GET, localhost, 4 clients, using ab + h2load
Influence of Parallel Requests:
Max 24 vs 384 balls in play...
Processing a h2 connection is vastly different in terms of (Connection/Request/Thread).
In h2, 1 connection has
n ongoing request, worked on by
Existing runtimes are optimized for 1-1-1
Pioneered by Matthew Steele and Bryan McQuade!
Donated by GSMA and greenbytes in July
Things to observe:
"You're off the edge of the map, mate. Here there be monsters."
Ericcson Research working on a Draft, greenbytes consulting