tests/e2e: concurrent UDP endpoints must share a reverse tunnel intact

Covers gost#911 at the wire level: four UDP source sockets send through one
RUDP reverse tunnel bound on a relay server, and every datagram must come
back byte-identical to its own endpoint.

A frame torn by an interleaved writer shows up twice — as a corrupted
payload, and as the torn frame leaving half a header in the stream, which the
far end reads as "unexpected EOF" and answers by rebuilding the whole tunnel.
So the suite asserts both: no endpoint sees a corrupted datagram, and no
endpoint sees a receive gap long enough to be a teardown and rebind.

Verified against a binary built before the fix (4 endpoints, 35% loss, 1.4s
gaps, suite fails) and after it (0% loss, no corruption, 0.2s gaps, passes).
This commit is contained in:
ginuerzh
2026-10-03 16:27:05 +08:00
parent e9fab95872
commit d7656ef3de
5 changed files with 346 additions and 0 deletions
+105
View File
@@ -0,0 +1,105 @@
"""Concurrent UDP sender for the gost#911 reverse-tunnel regression test.
Each socket is an independent source endpoint sharing ONE reverse tunnel
stream. Every datagram is a run of its own endpoint id, so a reply that does
not match the id byte exactly is a frame corrupted by an interleaved writer
rather than plain loss.
Usage: udp_rudp_concurrent.py <host> <port> <count> <duration_seconds>
"""
import socket
import sys
import threading
import time
def main():
host = sys.argv[1]
port = int(sys.argv[2])
count = int(sys.argv[3])
duration = float(sys.argv[4])
payload_len = 1200
end = time.monotonic() + duration
results = {}
lock = threading.Lock()
def endpoint(endpoint_id):
payload = bytes([endpoint_id]) * payload_len
sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
sock.bind(("0.0.0.0", 0))
sock.settimeout(0.2)
sent = received = damaged = 0
largest_gap = 0.0
last = time.monotonic()
while time.monotonic() < end:
try:
sock.sendto(payload, (host, port))
sent += 1
except OSError:
pass
while True:
try:
data, _ = sock.recvfrom(65535)
except socket.timeout:
break
except OSError:
break
received += 1
if data != payload:
damaged += 1
now = time.monotonic()
largest_gap = max(largest_gap, now - last)
last = now
largest_gap = max(largest_gap, time.monotonic() - last)
with lock:
results[endpoint_id] = (sent, received, damaged, round(largest_gap, 3))
sock.close()
threads = [
threading.Thread(target=endpoint, args=(i + 1,)) for i in range(count)
]
for t in threads:
t.start()
for t in threads:
t.join()
failures = []
for endpoint_id in sorted(results):
sent, received, damaged, gap = results[endpoint_id]
loss_pct = 100.0 * (1.0 - received / sent) if sent else 100.0
print(
f"endpoint {endpoint_id}: sent={sent} received={received} "
f"damaged={damaged} loss={loss_pct:.1f}% max_gap={gap}"
)
# Unsent datagrams are outside our control (the UDP socket's own
# buffer); anything that comes BACK must be byte-identical.
if damaged:
failures.append(
f"endpoint {endpoint_id}: {damaged} datagram(s) came back "
f"corrupted — frames interleaved on the shared tunnel stream"
)
# A torn frame makes the far end read io.ErrUnexpectedEOF, which
# rebuilds the whole reverse tunnel. That shows up as a receive gap of
# seconds, not milliseconds.
if gap >= 1.0:
failures.append(
f"endpoint {endpoint_id}: {gap}s receive gap — the reverse "
f"tunnel was torn down and rebound under active traffic"
)
if failures:
for f in failures:
print(f"FAIL: {f}", file=sys.stderr)
sys.exit(1)
print(f"PASS: {count} concurrent endpoints, no corruption, no tunnel reset")
sys.exit(0)
if __name__ == "__main__":
main()