chunkloris: node http2 (native)

part of the chunkloris per-chunk amplification survey. this page is the per-server record for node http2 (native) under http/2 (h2c) data frames.

at a glance

server: node http2 (native) node-22
runtime: node-22
ecosystem: node
concurrency model: event-loop
parser: nghttp2 (C, bundled)
delivery granularity: per-chunk
chunk-limit helper: none exposed by the framework
verdict: per-frame — the parser/dispatcher boundary delivers one event per protocol frame (h2 / h3 DATA frame, or ws data frame). cpu cost under paced mode b is measurable per frame.
scaling exponent (mode a): 0.86 (wall time vs N, log-log slope across common cells)
scaling exponent (mode b): 1.00

measurements

all cells run on a 1-vcpu docker container. cpu cost is derived from the target container’s cgroup v2 cpu.stat usage_usec delta around each cell.

mode	N	wall (s)	server cpu %	µs / frame	basis	ok
`A-h2-bridge`	50,000	0.204	57.6	2.348	server-cpu-cgroup	✓
`A-h2-bridge`	100,000	0.340	33.4	1.137	server-cpu-cgroup	✓
`A-h2-bridge`	250,000	0.805	25.4	0.818	server-cpu-cgroup	✓
`B-h2-paced-100us`	50,000	5.410	5.5	5.999	server-cpu-cgroup	✓
`B-h2-paced-100us`	100,000	10.771	4.8	5.164	server-cpu-cgroup	✓
`B-h2-paced-100us`	250,000	26.926	4.8	5.159	server-cpu-cgroup	✓

what this means

the parser/dispatcher path on this server delivers one event per protocol frame (a http/2 (h2c) data frames DATA frame or ws frame), so an attacker who sends a request body as N one-byte frames consumes roughly N × (mode-b µs/frame) of server cpu on a single core.

what to do today

if this is an h2 origin, prefer a frontend that terminates h2 into h1 with proxy_request_buffering on upstream.
consider imposing a per-stream DATA-frame credit (count, not bytes) before forwarding the body to the application handler.
HTTP/2 byte-level flow control (WINDOW_UPDATE) does not bound the number of frames; configure stream-frame-rate limits where the implementation exposes them.

reproducer

the full reproducer for this server is in the paper repo. the docker container pins node http2 (native) node-22 and constrains the test container to a single cpu (--cpus=1). the prober script implements mode a (bridge-coalesced) and mode b (paced 100 µs) per the methodology section.

see the draft pdf for the full per-framework discussion.