Files
PolyGateway/tests/unit/test_backpressure.py
T
iomgaa 45073486a7 fix: reparent governance backend failures under GatewayUnavailableError (issue #7)
A fail-closed limiter or breaker backend means the scope cannot emit a
single request, which is exactly scope-level unavailability. But the error
sat directly under PolyGatewayError, so a caller writing only
`except GatewayUnavailableError` dropped it into the catch-all branch:
Redis blips once and a backlog of tasks burns its business failure budget
into the dead letter queue, over a fault a restart would clear.

Three gate paths leak to callers rather than being absorbed by
_record_quietly (try_acquire, try_enter, progress_age_s); each is now
pinned by a test, since none of them had one before.

The two unknown-source sites move to SourceNotConfiguredError instead of
following along. They report a misconfigured source name, not an outage,
and letting them into the retryable family would be the mirror of the bug
being fixed here: the task would retry forever and never surface.
2026-08-06 04:53:52 -04:00

329 lines
13 KiB
Python

"""背压 stall 双条件判死 + poll jitter + 记账侧降级(M2 设计 §4/§10)。
保真蓝本 CHS governance.py:200-285: local_waited(本地 monotonic,调用级
累计不重置)与 progress_age(全局活性)**同时**超 stall_window 才判死;
poll 间隔带 [0.5p, 1.0p] jitter 防惊群。
"""
import asyncio
import pytest
from polygateway.backends.memory.breaker import InMemoryGate
from polygateway.backends.memory.limiter import InMemoryLimiter
from polygateway.errors import (
AllSourcesExhausted,
GatewayUnavailableError,
GovernanceBackendError,
SourceNotConfiguredError,
TransientError,
)
from polygateway.middleware.retry import RetryMW, backoff_delay
from polygateway.sources import RoundRobinSelector, SourceCooldownMemo
from polygateway.types import (
BackpressurePolicy,
BreakerConfig,
ChatRequest,
GlobalLimits,
RetryPolicy,
)
from tests.contracts.conftest import FakeClock, make_source
from tests.unit.test_retry import FakeTransport, _ok
_BREAKER = BreakerConfig(fail_threshold=3, cooldown_s=60.0, probe_ttl_s=120.0)
_NO_GLOBAL = GlobalLimits(max_concurrency=0, rpm=0, tpm=0)
_REQ = ChatRequest(messages=[{"role": "user", "content": "hi"}])
_STALL = 300.0
class BoundedSleep:
"""每次 sleep 执行注入的副作用;超过上限仍在轮询 = 判死逻辑失效,炸出而非死循环。"""
def __init__(self, side_effect=None, limit: int = 10):
self.delays: list[float] = []
self._side_effect = side_effect
self._limit = limit
async def __call__(self, seconds: float) -> None:
self.delays.append(seconds)
if len(self.delays) > self._limit:
raise RuntimeError(f"超过 {self._limit} 次轮询仍未判死/未获 permit")
if self._side_effect is not None:
await self._side_effect(len(self.delays))
def _mw(sources, limiter, script, *, clock, sleep, rng=lambda: 0.0, quota_full="wait", gate=None):
return RetryMW(
scope="llm",
sources=sources,
selector=RoundRobinSelector(),
limiter=limiter,
gate=gate or InMemoryGate(config=_BREAKER, now=clock),
transport=FakeTransport(script),
retry=RetryPolicy(max_attempts=3, backoff_base_s=2.0, backoff_max_s=30.0),
backpressure=BackpressurePolicy(stall_window_s=_STALL, poll_interval_s=0.01),
quota_full=quota_full,
cooldown_memo=SourceCooldownMemo(now=clock),
emitter=None,
now=clock,
sleep=sleep,
rng=rng,
)
def _blocked_limiter(clock):
"""单并发源被外部占满 → RetryMW 进入 wait 轮询分支。"""
src = make_source(max_concurrency=1)
limiter = InMemoryLimiter(
scope="llm",
sources={"s1": src},
global_limits=_NO_GLOBAL,
lease_ttl_s=10_000.0,
now=clock,
)
return src, limiter
class TestStallQuadrants:
async def test_both_windows_exceeded_raises_stalled(self):
"""双超窗: 本地等待与全局无进展同时 > stall_window → stalled。"""
clock = FakeClock()
src, limiter = _blocked_limiter(clock)
_held = await limiter.try_acquire("s1", 0)
async def advance(_n):
clock.advance(_STALL + 100)
mw = _mw([src], limiter, [], clock=clock, sleep=BoundedSleep(advance))
with pytest.raises(AllSourcesExhausted) as ei:
await mw(_REQ)
assert ei.value.reason == "stalled"
assert ei.value.per_source_reasons == {"s1": "rate_limited"}
async def test_local_exceeded_but_global_fresh_keeps_waiting(self):
"""仅本地超窗: 别人一直在出餐 → 不判死,等到 permit 后正常成功。"""
clock = FakeClock()
src, limiter = _blocked_limiter(clock)
held = await limiter.try_acquire("s1", 0)
async def advance_and_feed(n):
clock.advance(_STALL + 100) # 本地远超窗
await limiter.mark_progress() # 但全局刚出过餐
if n >= 3:
await held.release() # 第 3 轮让出 permit
mw = _mw([src], limiter, [_ok()], clock=clock, sleep=BoundedSleep(advance_and_feed))
resp = await mw(_REQ)
assert resp.content == "ok"
async def test_global_stale_but_local_fresh_keeps_waiting(self):
"""仅全局超窗(从未出餐 age=inf): 本地才刚开始等 → 不判死。"""
clock = FakeClock()
src, limiter = _blocked_limiter(clock)
held = await limiter.try_acquire("s1", 0)
async def release_soon(n):
if n >= 2: # 本地累计 poll 极短,未超窗
await held.release()
mw = _mw([src], limiter, [_ok()], clock=clock, sleep=BoundedSleep(release_soon))
resp = await mw(_REQ)
assert resp.content == "ok"
async def test_poll_jitter_bounds(self):
"""wait 轮询间隔 ∈ [0.5p, 1.0p](CHS governance.py:283-285)。"""
clock = FakeClock()
src, limiter = _blocked_limiter(clock)
held = await limiter.try_acquire("s1", 0)
async def release_soon(n):
if n >= 2:
await held.release()
for rng_v, expected in ((0.0, 0.005), (1.0, 0.01)):
clock2 = FakeClock()
src2, limiter2 = _blocked_limiter(clock2)
held2 = await limiter2.try_acquire("s1", 0)
async def release2(n, _h=held2):
if n >= 2:
await _h.release()
sleep = BoundedSleep(release2)
mw = _mw([src2], limiter2, [_ok()], clock=clock2, sleep=sleep, rng=lambda v=rng_v: v)
await mw(_REQ)
assert sleep.delays[0] == pytest.approx(expected)
async def test_fail_fast_unaffected(self):
"""fail_fast 路径回归: 不进入 stall 判定,立即 quota_exhausted。"""
clock = FakeClock()
src, limiter = _blocked_limiter(clock)
_held = await limiter.try_acquire("s1", 0)
mw = _mw([src], limiter, [], clock=clock, sleep=BoundedSleep(), quota_full="fail_fast")
with pytest.raises(AllSourcesExhausted) as ei:
await mw(_REQ)
assert ei.value.reason == "quota_exhausted"
async def test_cancellation_pierces_wait_loop(self):
"""stall 等待中的取消穿透(铁律): sleep 可取消,任务立即终止。"""
clock = FakeClock()
src, limiter = _blocked_limiter(clock)
_held = await limiter.try_acquire("s1", 0)
mw = _mw([src], limiter, [], clock=clock, sleep=asyncio.sleep)
task = asyncio.create_task(mw(_REQ))
await asyncio.sleep(0.03)
task.cancel()
with pytest.raises(asyncio.CancelledError):
await task
class _GateSuccessBroken(InMemoryGate):
async def record_success(self, entry):
raise GovernanceBackendError("redis 抖动", scope="llm")
class _GateFailureBroken(InMemoryGate):
async def record_failure(self, entry, reason, force_open):
raise GovernanceBackendError("redis 抖动", scope="llm")
class _LimiterProgressBroken(InMemoryLimiter):
async def mark_progress(self):
raise GovernanceBackendError("redis 抖动", scope="llm")
class TestAccountingDegradation:
"""记账侧降级(设计 §10,ARCH §7.3 勘误): 调用已完成,写回失败不冒泡。"""
async def test_record_success_failure_does_not_lose_response(self):
clock = FakeClock()
src = make_source()
limiter = InMemoryLimiter(
scope="llm", sources={"s1": src}, global_limits=_NO_GLOBAL, now=clock
)
gate = _GateSuccessBroken(config=_BREAKER, now=clock)
mw = _mw([src], limiter, [_ok()], clock=clock, sleep=BoundedSleep(), gate=gate)
resp = await mw(_REQ)
assert resp.content == "ok" # 真实成功响应不因记账失败被丢弃
async def test_mark_progress_failure_does_not_lose_response(self):
clock = FakeClock()
src = make_source()
limiter = _LimiterProgressBroken(
scope="llm", sources={"s1": src}, global_limits=_NO_GLOBAL, now=clock
)
mw = _mw([src], limiter, [_ok()], clock=clock, sleep=BoundedSleep())
resp = await mw(_REQ)
assert resp.content == "ok"
async def test_record_failure_failure_does_not_mask_retry(self):
"""失败记账挂掉 → 原始尝试异常不被掩盖,重试照常换发并成功。"""
clock = FakeClock()
src = make_source()
limiter = InMemoryLimiter(
scope="llm", sources={"s1": src}, global_limits=_NO_GLOBAL, now=clock
)
gate = _GateFailureBroken(config=_BREAKER, now=clock)
script = [TransientError("boom", status_code=500), _ok()]
mw = _mw([src], limiter, script, clock=clock, sleep=BoundedSleep(), gate=gate)
resp = await mw(_REQ)
assert resp.content == "ok"
class TestBackoffExtraction:
"""T5 提取的模块级纯函数(T9 Embedding 复用)。"""
def test_formula_matches_policy(self):
policy = RetryPolicy(max_attempts=3, backoff_base_s=2.0, backoff_max_s=30.0)
exc = TransientError("x", status_code=500)
assert backoff_delay(policy, 1, exc, lambda: 0.0) == pytest.approx(1.0) # 2*0.5
assert backoff_delay(policy, 1, exc, lambda: 1.0) == pytest.approx(3.0) # 2*1.5
assert backoff_delay(policy, 10, exc, lambda: 1.0) == pytest.approx(45.0) # 封顶 30*1.5
def test_retry_after_hint_wins_when_larger(self):
policy = RetryPolicy(max_attempts=3, backoff_base_s=2.0, backoff_max_s=30.0)
exc = TransientError("x", status_code=429, retry_after_s=7.5)
assert backoff_delay(policy, 1, exc, lambda: 0.0) == pytest.approx(7.5)
class TestQuotaGateProgressAge:
async def test_passthrough_and_wrap(self):
from polygateway.middleware.ratelimit import QuotaGate
class _L:
async def progress_age_s(self):
return 12.5
class _Broken:
async def progress_age_s(self):
raise OSError("down")
assert await QuotaGate(_L(), scope="llm").progress_age_s() == 12.5
with pytest.raises(GovernanceBackendError):
await QuotaGate(_Broken(), scope="llm").progress_age_s()
class TestUnknownSourceIsAssemblyDefect:
"""未知源 = 限流后端的源名单与治理循环对不上,是装配缺陷不是后端故障。
两个后端行为必须一致(Redis 版对应用例在 `test_redis_key_layout.py::
TestConversions::test_unknown_source_rejected`);内存版此前无覆盖,
该分支从未被测过(issue #7 §3.4)。
"""
def test_memory_limiter_rejects_unknown_source(self):
limiter = InMemoryLimiter(
scope="llm", sources={"s1": make_source("s1")}, global_limits=_NO_GLOBAL
)
with pytest.raises(SourceNotConfiguredError) as ei:
limiter._cfg("nope")
# 关键: 若归入 scope 级家族,配置写错的任务会永远延期重投、永不进死信
assert not isinstance(ei.value, GatewayUnavailableError)
class TestGateFailuresReachCallersAsScopeLevel:
"""三条闸门泄漏路径必须以 scope 级不可用的形态到达调用方(issue #7)。
记账路径由 `_record_quietly` 降级为 warning,但闸门路径没有那层包裹,会一路
抛给调用方。只写 `except GatewayUnavailableError` 的调用方此前接不住,后果
是 Redis 抖一下就让积压任务烧掉业务失败预算进死信——而那是运维重启即可恢复
的故障。三条路径逐一钉住,防止将来任何一条被漏掉。
"""
async def test_try_acquire_failure_is_scope_level(self):
from polygateway.middleware.ratelimit import QuotaGate
class _Broken:
async def try_acquire(self, name, est):
raise OSError("down")
with pytest.raises(GatewayUnavailableError) as ei:
await QuotaGate(_Broken(), scope="LLM").try_acquire(make_source("s1"))
assert ei.value.scope == "llm"
assert ei.value.reason == "governance_backend_down"
assert ei.value.retry_after_s > 0 # 0 会让积压任务零延迟冲击已挂的后端
async def test_try_enter_failure_is_scope_level(self):
from polygateway.middleware.breaker import BreakerGate
class _Broken:
async def try_enter(self, name, owner):
raise OSError("down")
with pytest.raises(GatewayUnavailableError) as ei:
await BreakerGate(_Broken(), scope="LLM").try_enter(make_source("s1"), "owner")
assert ei.value.scope == "llm"
assert ei.value.reason == "governance_backend_down"
async def test_progress_age_failure_is_scope_level(self):
from polygateway.middleware.ratelimit import QuotaGate
class _Broken:
async def progress_age_s(self):
raise OSError("down")
with pytest.raises(GatewayUnavailableError) as ei:
await QuotaGate(_Broken(), scope="LLM").progress_age_s()
assert ei.value.scope == "llm"
assert ei.value.reason == "governance_backend_down"