File size: 21,361 Bytes
cc036ff | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 | """
Conftest for concurrent operations tests
This module provides fixtures for testing race conditions, deadlocks,
and resource cleanup issues that only manifest under concurrent access.
Fixtures are organized into:
1. Threading fixtures for cache tests
2. Asyncio fixtures for async tests
3. Database fixtures for lock tests
4. Resource tracking fixtures
IMPORTANT: SQLite Concurrency Limitations
-------------------------------------------
SQLite has limited concurrent write support:
- Only one writer at a time (serialized access)
- Multiple readers allowed (WITH one writer or zero writers)
- For true concurrency, use PostgreSQL with SERIALIZABLE isolation
Tests in this suite focus on:
- Read-heavy concurrent operations (SQLite can handle)
- Thread-safe cache access (in-memory, no DB locking)
- Async operation coordination (event loop concurrency)
- Documented PostgreSQL behavior for true parallel writes
For production deployment with PostgreSQL:
- Deadlocks can occur with conflicting lock orders
- SERIALIZABLE isolation prevents phantom reads
- Connection pool exhaustion under high load
- SELECT FOR UPDATE for pessimistic locking
"""
import asyncio
import gc
import os
import pytest
import tempfile
import threading
import time
import uuid
from collections import OrderedDict
from contextlib import contextmanager
from datetime import datetime, timedelta
from typing import Any, Dict, List, Optional
from unittest.mock import MagicMock, patch
from sqlalchemy import create_engine
from sqlalchemy.orm import Session, sessionmaker
from core.governance_cache import GovernanceCache
from core.database import Base
# ============================================================================
# Threading Fixtures for Cache Tests
# ============================================================================
@pytest.fixture(scope="function")
def concurrent_cache():
"""
Create a GovernanceCache instance for concurrent testing.
Thread-safe by design (threading.Lock protects internal state).
Tests will verify lock contention behavior and cache consistency.
Usage:
def test_cache_concurrent_write(concurrent_cache):
cache = concurrent_cache(max_size=100, thread_count=10)
# Launch 10 threads writing to cache
# Verify no data corruption
"""
def _create_cache(max_size: int = 1000, ttl_seconds: int = 60):
"""Create cache with specified parameters."""
return GovernanceCache(max_size=max_size, ttl_seconds=ttl_seconds)
return _create_cache
@pytest.fixture(scope="function")
def assert_cache_consistency():
"""
Verify cache consistency after concurrent operations.
Checks that:
- No data loss (all writes preserved)
- No corrupted entries (all values valid)
- Cache size within bounds (max_size respected)
- Statistics accurate (hits/misses/evictions)
Usage:
assert_cache_consistency(cache, expected_entries=100)
"""
def _assert_consistency(
cache: GovernanceCache,
expected_entries: Optional[int] = None,
max_size: Optional[int] = None
):
"""Assert cache is consistent after concurrent operations."""
stats = cache.get_stats()
# Verify cache size
if max_size:
assert stats["size"] <= max_size, f"Cache size {stats['size']} exceeds max {max_size}"
# Verify statistics are non-negative
assert stats["hits"] >= 0
assert stats["misses"] >= 0
assert stats["evictions"] >= 0
assert stats["invalidations"] >= 0
# Verify hit rate is valid
assert 0 <= stats["hit_rate"] <= 100
# Verify expected entries (with tolerance for evictions)
if expected_entries:
# Allow 10% tolerance for LRU eviction during writes
min_expected = int(expected_entries * 0.9)
assert stats["size"] >= min_expected, \
f"Cache size {stats['size']} below expected {min_expected}"
return _assert_consistency
@pytest.fixture(scope="function")
def timed_operation():
"""
Measure execution time for contention detection.
Returns operation duration in milliseconds. Useful for detecting
lock contention (operations taking >50ms under high contention).
Usage:
duration_ms = timed_operation(lambda: cache.get("agent", "action"))
assert duration_ms < 50, "Lock contention detected"
"""
def _time_operation(operation_fn, *args, **kwargs) -> float:
"""
Execute operation and return duration in milliseconds.
Args:
operation_fn: Function to execute
*args, **kwargs: Arguments passed to operation_fn
Returns:
Duration in milliseconds
"""
start_time = time.perf_counter()
result = operation_fn(*args, **kwargs)
end_time = time.perf_counter()
duration_ms = (end_time - start_time) * 1000
return duration_ms
return _time_operation
# ============================================================================
# Asyncio Fixtures for Async Tests
# ============================================================================
@pytest.fixture(scope="function")
def run_async_tasks():
"""
Run N async tasks concurrently using asyncio.gather.
Forces true concurrent execution (not sequential) for async race
condition testing. Verifies no state leakage between tasks.
Usage:
async def my_task(task_id):
return await service.create_episode(f"session_{task_id}")
results = await run_async_tasks(my_task, count=10)
assert len(results) == 10
"""
async def _run_tasks(coro, count: int, **kwargs) -> List[Any]:
"""
Run async coroutine concurrently N times.
Args:
coro: Async coroutine function (or lambda returning coroutine)
count: Number of concurrent tasks to run
**kwargs: Arguments passed to each coroutine
Returns:
List of results from all tasks
"""
tasks = [coro(i, **kwargs) if asyncio.iscoroutinefunction(coro) else coro for i in range(count)]
results = await asyncio.gather(*tasks, return_exceptions=True)
return results
return _run_tasks
@pytest.fixture(scope="function")
def assert_no_duplicate_ids():
"""
Verify no duplicate IDs in concurrent creation results.
Common bug: Race condition in ID generation causes collisions.
This fixture detects duplicate IDs from concurrent operations.
Usage:
results = await run_async_tasks(create_episode, count=10)
episode_ids = [r.id for r in results]
assert_no_duplicate_ids(episode_ids)
"""
def _assert_no_duplicates(items: List[Any], id_extractor=None):
"""
Assert no duplicate IDs in items.
Args:
items: List of items (objects or IDs)
id_extractor: Optional function to extract ID from item
Raises:
AssertionError: If duplicate IDs found
"""
if id_extractor:
ids = [id_extractor(item) for item in items]
elif isinstance(items[0], (str, int)):
ids = items
else:
# Assume objects with .id attribute
ids = [item.id for item in items]
unique_ids = set(ids)
if len(ids) != len(unique_ids):
# Find duplicates
seen = set()
duplicates = [x for x in ids if x in seen or seen.add(x)]
raise AssertionError(f"Duplicate IDs found: {duplicates}")
return _assert_no_duplicates
# ============================================================================
# Database Fixtures for Lock Tests
# ============================================================================
@pytest.fixture(scope="function")
def two_agent_sessions():
"""
Create two database sessions for deadlock testing.
Simulates two concurrent transactions that might deadlock when
updating rows in different orders. Each session has independent
transaction context.
Usage:
session1, session2 = two_agent_sessions()
# Transaction 1: Update agent1 then agent2
session1.query(Agent).filter(...).update(...)
# Transaction 2: Update agent2 then agent1 (reverse order)
session2.query(Agent).filter(...).update(...)
"""
# Create in-memory SQLite database
fd, db_path = tempfile.mkstemp(suffix='.db')
os.close(fd)
engine = create_engine(
f"sqlite:///{db_path}",
connect_args={"check_same_thread": False},
echo=False
)
# Create tables
Base.metadata.create_all(engine, checkfirst=True)
# Create two session factories
SessionLocal1 = sessionmaker(autocommit=False, autoflush=False, bind=engine)
SessionLocal2 = sessionmaker(autocommit=False, autoflush=False, bind=engine)
session1 = SessionLocal1()
session2 = SessionLocal2()
yield session1, session2
# Cleanup
session1.close()
session2.close()
engine.dispose()
try:
os.unlink(db_path)
except Exception:
pass
@pytest.fixture(scope="function")
def assert_transaction_rollback():
"""
Verify transaction was rolled back after deadlock/error.
Checks that database state matches expected after rollback.
Useful for verifying deadlock handling.
Usage:
# After deadlock detected
assert_transaction_rollback(session, expected_state={"agent1": 0.7})
"""
def _assert_rollback(session: Session, expected_state: Dict[str, Any]):
"""
Assert database state matches expected after rollback.
Args:
session: Database session to query
expected_state: Dict of expected values {"key": value}
"""
from core.models import AgentRegistry
for key, expected_value in expected_state.items():
if key.startswith("agent_"):
# Query agent by ID or name
agent = session.query(AgentRegistry).filter(
(AgentRegistry.id == key) | (AgentRegistry.name == key)
).first()
if agent:
actual_value = agent.confidence_score
assert actual_value == expected_value, \
f"Rollback failed: {key} has {actual_value}, expected {expected_value}"
return _assert_rollback
# ============================================================================
# Resource Tracking Fixtures
# ============================================================================
@pytest.fixture(scope="function")
def connection_counter():
"""
Track DB connection count for leak detection.
Monitors open database connections before/after operations
to detect connection leaks (connections not properly closed).
Usage:
counter = connection_counter()
before = counter.count()
# Run operations that might leak connections
await many_async_db_calls()
after = counter.count()
assert after <= before + 2, "Connections leaked!"
"""
class _ConnectionCounter:
def __init__(self):
self.engine = None
def count(self) -> int:
"""
Count open database connections.
For SQLite: Check open file handles
For PostgreSQL: Query pg_stat_activity
Returns:
Number of open connections
"""
# Simplified implementation for SQLite
# In production with PostgreSQL, query pg_stat_activity
import sqlite3
count = 0
for thread in threading.enumerate():
if hasattr(thread, '_connection'):
count += 1
return count
def detailed_count(self) -> Dict[str, int]:
"""Return detailed connection statistics."""
return {
"total": self.count(),
"active": len([t for t in threading.enumerate() if hasattr(t, '_connection')]),
}
return _ConnectionCounter()
@pytest.fixture(scope="function")
def leak_detector():
"""
Detect resource leaks (memory, file handles, connections).
Tracks object counts before/after operations to identify leaks.
Uses garbage collector for accurate memory tracking.
Usage:
detector = leak_detector("database_connection")
detector.start()
# Run operations that might leak
await risky_async_operations()
leaked = detector.stop()
assert leaked == 0, f"Leaked {leaked} resources"
"""
class _LeakDetector:
def __init__(self, resource_type: str):
self.resource_type = resource_type
self.before_count = 0
self.after_count = 0
def start(self):
"""Record initial resource count."""
gc.collect() # Force GC before counting
self.before_count = self._count_resources()
def stop(self) -> int:
"""
Record final resource count and return leak count.
Returns:
Number of leaked resources (0 if no leak)
"""
gc.collect() # Force GC before counting
self.after_count = self._count_resources()
return max(0, self.after_count - self.before_count)
def _count_resources(self) -> int:
"""Count resources of tracked type."""
if self.resource_type == "database_connection":
# Count open DB connections
return len([
obj for obj in gc.get_objects()
if hasattr(obj, '__class__') and
'Connection' in obj.__class__.__name__ and
hasattr(obj, 'connection')
])
elif self.resource_type == "thread":
# Count threads (excluding main thread)
return len([t for t in threading.enumerate() if t.name != 'MainThread'])
elif self.resource_type == "object":
# Count all Python objects (for memory leak detection)
return len(gc.get_objects())
else:
return 0
return _LeakDetector
@pytest.fixture(scope="function")
def get_object_count():
"""
Get current Python object count for memory leak testing.
Uses garbage collector to count all live Python objects.
Useful for detecting memory leaks in long-running operations.
Usage:
initial_objects = get_object_count()
# Run many operations
for i in range(1000):
await service.create_episode(...)
gc.collect()
final_objects = get_object_count()
assert final_objects - initial_objects < 100, "Memory leak detected"
"""
def _count_objects() -> int:
"""Count all Python objects in memory."""
gc.collect()
return len(gc.get_objects())
return _count_objects
# ============================================================================
# SQLite Concurrency Documentation Fixture
# ============================================================================
@pytest.fixture(scope="session", autouse=True)
def sqlite_concurrency_notes():
"""
Document SQLite concurrency limitations for all tests.
This fixture auto-documents test behavior expectations for
SQLite vs PostgreSQL concurrency differences.
SQLite Limitations:
- Only one writer at a time (serialized)
- Multiple readers allowed (WITH one writer)
- Write operations are queued (first-come, first-served)
- No true parallel write concurrency
PostgreSQL Advantages:
- Multiple writers with MVCC
- True parallel writes with row-level locking
- SERIALIZABLE isolation prevents phantom reads
- Deadlock detection and automatic rollback
Test Strategy:
- Focus on read-heavy concurrency (SQLite friendly)
- Test cache thread-safety (in-memory, no DB locking)
- Test async coordination (event loop, not DB locking)
- Document expected PostgreSQL behavior for write-heavy tests
"""
# This is a documentation fixture - no action needed
# The docstring serves as inline documentation for test developers
pass
# ============================================================================
# Test Helpers
# ============================================================================
@pytest.fixture(scope="function")
def is_ci_environment():
"""
Detect if running in CI environment.
CI environments have different concurrency characteristics:
- Fewer CPU cores (slower concurrent execution)
- Shared resources (contention more likely)
- Timeouts may need adjustment
Usage:
if is_ci_environment():
timeout_seconds = 10
else:
timeout_seconds = 2
"""
return os.getenv("CI", "false").lower() == "true"
@pytest.fixture(scope="function")
def retry_on_deadlock():
"""
Retry function call on database deadlock.
Deadlocks are transient errors - retrying with backoff
is the standard handling strategy.
Usage:
result = retry_on_deadlock(
lambda: update_agent(session, agent_id, data),
max_retries=3
)
"""
def _retry(func, max_retries: int = 3, backoff_ms: int = 100):
"""
Retry function on deadlock/lock error.
Args:
func: Function to execute
max_retries: Maximum retry attempts
backoff_ms: Backoff delay between retries
Returns:
Function result
Raises:
Exception: If all retries exhausted
"""
last_exception = None
for attempt in range(max_retries):
try:
return func()
except Exception as e:
last_exception = e
error_msg = str(e).lower()
if "deadlock" in error_msg or "lock" in error_msg:
if attempt < max_retries - 1:
time.sleep(backoff_ms / 1000 * (attempt + 1))
continue
raise
raise last_exception
return _retry
# ============================================================================
# Performance Benchmarks
# ============================================================================
@pytest.fixture(scope="function")
def benchmark_concurrent_operations():
"""
Benchmark concurrent operation performance.
Measures throughput (ops/sec) and latency under various
concurrency levels to identify performance regressions.
Usage:
results = benchmark_concurrent_operations(
operation=lambda: cache.get("agent", "action"),
thread_count=50,
operations_per_thread=100
)
print(f"Throughput: {results['ops_per_sec']:.0f} ops/sec")
print(f"Latency P50: {results['latency_p50_ms']:.2f} ms")
"""
def _benchmark(
operation_fn,
thread_count: int = 10,
operations_per_thread: int = 100
) -> Dict[str, Any]:
"""
Benchmark operation under concurrent load.
Args:
operation_fn: Function to benchmark
thread_count: Number of concurrent threads
operations_per_thread: Operations per thread
Returns:
Dict with ops_per_sec, latency_p50_ms, latency_p99_ms
"""
errors = []
latencies = []
stop_event = threading.Event()
def worker(thread_id: int):
"""Worker thread for benchmark."""
thread_latencies = []
for i in range(operations_per_thread):
if stop_event.is_set():
break
start = time.perf_counter()
try:
operation_fn()
latency_ms = (time.perf_counter() - start) * 1000
thread_latencies.append(latency_ms)
except Exception as e:
errors.append(e)
stop_event.set()
break
latencies.extend(thread_latencies)
# Launch threads
start_time = time.perf_counter()
threads = [
threading.Thread(target=worker, args=(i,))
for i in range(thread_count)
]
for t in threads:
t.start()
for t in threads:
t.join()
end_time = time.perf_counter()
# Calculate metrics
total_duration = end_time - start_time
total_ops = len(latencies)
ops_per_sec = total_ops / total_duration if total_duration > 0 else 0
latencies_sorted = sorted(latencies)
p50_idx = int(len(latencies_sorted) * 0.5)
p99_idx = int(len(latencies_sorted) * 0.99)
return {
"ops_per_sec": ops_per_sec,
"latency_p50_ms": latencies_sorted[p50_idx] if latencies_sorted else 0,
"latency_p99_ms": latencies_sorted[p99_idx] if latencies_sorted else 0,
"errors": errors,
"total_operations": total_ops,
}
return _benchmark
|