mirror of
https://gitee.com/xmz7891/ykxmz.git
synced 2026-10-19 10:40:44 +00:00
修局域网/中转对比基准：它自己有两个 bug，害产品背了黑锅
这个基准一直是红的（中转那条路报「连接阶段失败: timed out」），
而且报告里写着"这是中转链路的产品 bug"。**查下来不是产品的问题。**
== bug 1：两条路用了同一个配对码 ==
基准为了排除互相干扰，给两条路**各起一台全新被控端**（设计是对的），
但两台都用了写死的 code="246813"。
中转有「一个配对码只属于一台设备」的规则：
第 1 台（跑局域网）注册成功
-> stop_agent 杀掉进程，但**中转还以为它在线**（last_seen 没过期）
第 2 台（跑中转）用同一个码注册 -> **被拒**
中转日志：配对码 246813 已被 ... 占用，拒绝 ...
而 start_agent 只检查"有没有设备在线" —— 第 1 台僵尸还挂着，
于是**以为新的一台登记成功了**
-> 没有任何进程在回包 -> 超时
对照实验（新增 _build/verify_code_reuse.py）：
场景 A 同一个码 -> **不通**
场景 B 各用不同码 -> 通
修法：两条路各用不同码；start_agent 的判据从"有没有在线设备"
改成"有没有**配对码正好是这一台要用的码**的在线设备"。
== bug 2：attach 用的还是旧码（我第一版补丁也漏了这处）==
relay_connect() 里 `client.attach(code)` 用的还是局域网那个码 ——
attach 到一台已经死掉、但中转以为在线的设备上，pair 自然没人回。
修成 `attach(relay_code)`。
== 顺带更新一段过时注释 ==
原文写着「当前版本经中转只稳定支撑前 1~2 次交互……这是中转链路的
产品 bug」，并据此把中转的轮数调小、超时调短。
那个 bug **已经修好了**（根因：一条 HTTP/1.1 连接被长轮询和发命令
两个线程同时用，撞状态机抛 CannotSendRequest("Idle")，
异常冒到 _client_loop 外面把 client-relay 线程打死）。
注释不更新会误导后人以为产品还有这个问题。
== 修好之后的实测（本地中转，30 次/路）==
局域网 握手 16ms 成功 30/30 (100%) 中位 0ms P90 0ms 最大 1ms
中转 握手 46ms 成功 30/30 (100%) 中位 15ms P90 29ms 最大 32ms
这是本地中转，量的是**协议本身的开销**（长轮询往返 vs 直连 TCP）；
真实跨网络那条腿的 1.0~1.5 秒来自隧道和 4G，是另一回事。
结论：局域网确实是"毫秒级、不耗流量"的那条根本路径 ——
中位 0ms vs 15ms，而且不走任何外网。
踩坑记录：这两处都是"看起来像产品 bug、实际是测试自己的 bug"。
先怀疑测量工具，比先怀疑产品更省时间 ——
上一次（管理页缓存头）也是我拿 curl -I 测 GET 白查了半天。
This commit is contained in:
xmz7891
@@ -0,0 +1,241 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
r'''复刻 bench_lan_vs_relay 的中转那一半，抓被控端日志。
|
||||
|
||||
我自己那个 ins_relay_flow.py 是中转 4/4 通过的，
|
||||
但 bench_lan_vs_relay.py 的中转那半报"连接阶段失败: timed out"。
|
||||
差异一定在设置里，所以要逐项对齐 bench 的写法：
|
||||
* RELAY_POLL=2（我那边是 5）
|
||||
* 用 YKXMZ_DATA 指到临时目录 + --silent（我那边直接写 windows/data）
|
||||
* 每条路起**一台全新的**被控端
|
||||
* 先 attach 再 pair（bench 的 relay_connect 就是这么做的）
|
||||
|
||||
把每一步的耗时和被控端日志都打出来。
|
||||
'''
|
||||
from __future__ import annotations
|
||||
|
||||
import json
|
||||
import os
|
||||
import socket
|
||||
import subprocess
|
||||
import sys
|
||||
import tempfile
|
||||
import time
|
||||
import urllib.error
|
||||
import urllib.request
|
||||
from pathlib import Path
|
||||
|
||||
ROOT = Path(__file__).resolve().parent.parent
|
||||
WINDOWS = ROOT / "windows"
|
||||
UA = {"User-Agent": "LanRemote/1.8.2"}
|
||||
CODE = "975310"
|
||||
|
||||
|
||||
def free_port() -> int:
|
||||
s = socket.socket()
|
||||
s.bind(("127.0.0.1", 0))
|
||||
p = s.getsockname()[1]
|
||||
s.close()
|
||||
return p
|
||||
|
||||
|
||||
def post(base, path, body=b"", headers=None, timeout=20.0):
|
||||
h = dict(UA)
|
||||
h.update(headers or {})
|
||||
req = urllib.request.Request(base + path, data=body, headers=h,
|
||||
method="POST")
|
||||
try:
|
||||
with urllib.request.urlopen(req, timeout=timeout) as resp:
|
||||
return resp.status, resp.read()
|
||||
except urllib.error.HTTPError as exc:
|
||||
return exc.code, exc.read()
|
||||
except Exception as exc: # noqa: BLE001
|
||||
return 0, str(exc).encode()
|
||||
|
||||
|
||||
def get(base, path, timeout=5.0):
|
||||
req = urllib.request.Request(base + path, headers=UA)
|
||||
try:
|
||||
with urllib.request.urlopen(req, timeout=timeout) as resp:
|
||||
return resp.status, resp.read()
|
||||
except Exception as exc: # noqa: BLE001
|
||||
return 0, str(exc).encode()
|
||||
|
||||
|
||||
def main() -> int:
|
||||
work = Path(tempfile.mkdtemp(prefix="benchdiag_"))
|
||||
relay_port = free_port()
|
||||
agent_port = free_port()
|
||||
base = "http://127.0.0.1:%d" % relay_port
|
||||
relay_proc = agent_proc = None
|
||||
relay_log = agent_log = None
|
||||
try:
|
||||
print(" 中转端口 %d 被控端端口 %d" % (relay_port, agent_port))
|
||||
|
||||
# ---- 1. 本地中转（对齐 bench：RELAY_POLL=2）----
|
||||
rdata = work / "r"
|
||||
rdata.mkdir(parents=True, exist_ok=True)
|
||||
env = dict(os.environ)
|
||||
env.update({"RELAY_PORT": str(relay_port), "RELAY_BIND": "127.0.0.1",
|
||||
"RELAY_DATA": str(rdata), "RELAY_POLL": "2",
|
||||
"PYTHONIOENCODING": "utf-8", "PYTHONUNBUFFERED": "1"})
|
||||
relay_log = open(work / "relay.log", "wb")
|
||||
relay_proc = subprocess.Popen(
|
||||
[sys.executable, "-u", str(ROOT / "relay" / "relay.py")], env=env,
|
||||
stdout=relay_log, stderr=subprocess.STDOUT)
|
||||
for _ in range(60):
|
||||
time.sleep(0.5)
|
||||
if get(base, "/v1/health")[0] == 200:
|
||||
break
|
||||
print(" 中转就绪")
|
||||
|
||||
# ---- 2. 被控端（对齐 bench：YKXMZ_DATA + --silent + 独立锁端口）----
|
||||
sys.path.insert(0, str(WINDOWS))
|
||||
from lanremote.config import Store
|
||||
adata = work / "a"
|
||||
adata.mkdir(parents=True, exist_ok=True)
|
||||
store = Store(adata)
|
||||
store.set_pairing_code(CODE)
|
||||
store.set("setup_done", True)
|
||||
store.set("allow_new_pairing", True)
|
||||
store.set("close_pair_after_first", False)
|
||||
store.set("server_name", "基准测试电脑")
|
||||
store.set("relay_url", base)
|
||||
store.set("port", agent_port)
|
||||
|
||||
aenv = dict(os.environ)
|
||||
aenv.update({"YKXMZ_DATA": str(adata),
|
||||
"YKXMZ_LOCK_PORT": str(free_port()),
|
||||
"PYTHONIOENCODING": "utf-8", "PYTHONUNBUFFERED": "1"})
|
||||
agent_log = open(work / "agent.log", "wb")
|
||||
agent_proc = subprocess.Popen(
|
||||
[sys.executable, "-u", str(WINDOWS / "run_agent.py"),
|
||||
"--silent", "--port", str(agent_port)],
|
||||
cwd=str(WINDOWS), env=aenv,
|
||||
stdout=agent_log, stderr=subprocess.STDOUT)
|
||||
|
||||
online = False
|
||||
for _ in range(45):
|
||||
time.sleep(1.0)
|
||||
if agent_proc.poll() is not None:
|
||||
print(" ✗ 被控端启动即退出，返回码 %s" % agent_proc.returncode)
|
||||
break
|
||||
st, body = get(base, "/v1/health", timeout=4)
|
||||
if st == 200:
|
||||
snap = json.loads(body.decode())
|
||||
if any(d.get("online") for d in snap.get("devices", [])):
|
||||
online = True
|
||||
break
|
||||
print(" 被控端登记: %s" % ("成功" if online else "**失败**"))
|
||||
if not online:
|
||||
return 1
|
||||
|
||||
# ---- 3. 完全照 bench 的 relay_connect 做 ----
|
||||
print()
|
||||
print(" ── 照 bench 的顺序：attach -> pair ──")
|
||||
t0 = time.time()
|
||||
st, body = post(base, "/v1/client/attach", json.dumps(
|
||||
{"code": CODE, "name": "基准测试手表"}).encode(),
|
||||
{"Content-Type": "application/json"})
|
||||
att = json.loads(body.decode())
|
||||
session = att.get("session", "")
|
||||
print(" attach: %dms ok=%s session=%s"
|
||||
% (int((time.time() - t0) * 1000), att.get("ok"),
|
||||
session[:8] if session else "无"))
|
||||
|
||||
payload = (json.dumps({"cmd": "pair", "deviceId": "bench-relay",
|
||||
"deviceName": "基准测试手表",
|
||||
"code": CODE}, ensure_ascii=False)
|
||||
+ "\n").encode()
|
||||
t0 = time.time()
|
||||
st, _ = post(base, "/v1/client/send", payload, {"X-Session": session})
|
||||
print(" send pair: %dms status=%s" % (int((time.time() - t0) * 1000), st))
|
||||
|
||||
# 轮询等回复，每轮都打点
|
||||
buf = b""
|
||||
deadline = time.time() + 40
|
||||
got = None
|
||||
round_no = 0
|
||||
while time.time() < deadline:
|
||||
round_no += 1
|
||||
remaining = deadline - time.time()
|
||||
t0 = time.time()
|
||||
st, chunk = post(base, "/v1/client/poll", b"",
|
||||
{"X-Session": session}, min(10.0, remaining))
|
||||
buf += chunk
|
||||
print(" poll#%d: %dms status=%s 收到 %d 字节"
|
||||
% (round_no, int((time.time() - t0) * 1000), st, len(chunk)))
|
||||
while b"\n" in buf:
|
||||
line, _, buf = buf.partition(b"\n")
|
||||
if not line.strip():
|
||||
continue
|
||||
msg = json.loads(line.decode())
|
||||
if "event" in msg:
|
||||
continue
|
||||
got = msg
|
||||
break
|
||||
if got is not None:
|
||||
break
|
||||
|
||||
if got is None:
|
||||
print(" ✗ 40 秒内没等到 pair 的回复")
|
||||
else:
|
||||
print(" ✓ pair 回复: %s" % json.dumps(got, ensure_ascii=False)[:120])
|
||||
|
||||
# ---- 4. 再来几次 ping ----
|
||||
if got is not None:
|
||||
for i in range(5):
|
||||
payload = (json.dumps({"cmd": "ping"}) + "\n").encode()
|
||||
post(base, "/v1/client/send", payload, {"X-Session": session})
|
||||
deadline = time.time() + 15
|
||||
ok = False
|
||||
while time.time() < deadline:
|
||||
st, chunk = post(base, "/v1/client/poll", b"",
|
||||
{"X-Session": session}, 8.0)
|
||||
buf += chunk
|
||||
while b"\n" in buf:
|
||||
line, _, buf = buf.partition(b"\n")
|
||||
if not line.strip():
|
||||
continue
|
||||
msg = json.loads(line.decode())
|
||||
if "event" in msg:
|
||||
continue
|
||||
ok = True
|
||||
break
|
||||
if ok:
|
||||
break
|
||||
print(" ping#%d: %s" % (i + 1, "OK" if ok else "**超时**"))
|
||||
|
||||
# ---- 5. 被控端日志 ----
|
||||
print()
|
||||
print(" ════ 被控端日志（最后 25 行）════")
|
||||
agent_log.flush()
|
||||
for line in (work / "agent.log").read_text(
|
||||
encoding="utf-8", errors="replace").splitlines()[-25:]:
|
||||
print(" " + line[:120])
|
||||
print()
|
||||
print(" ════ 进程还活着吗 ════")
|
||||
print(" agent poll = %s" % agent_proc.poll())
|
||||
print()
|
||||
print(" ════ 中转日志（最后 10 行）════")
|
||||
relay_log.flush()
|
||||
for line in (work / "relay.log").read_text(
|
||||
encoding="utf-8", errors="replace").splitlines()[-10:]:
|
||||
print(" " + line[:120])
|
||||
return 0
|
||||
finally:
|
||||
for p in (agent_proc, relay_proc):
|
||||
if p is not None:
|
||||
try:
|
||||
p.terminate()
|
||||
p.wait(timeout=6)
|
||||
except Exception: # noqa: BLE001
|
||||
try:
|
||||
p.kill()
|
||||
except Exception: # noqa: BLE001
|
||||
pass
|
||||
import shutil
|
||||
shutil.rmtree(work, ignore_errors=True)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
sys.exit(main())
|
||||
Reference in New Issue
Block a user