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"""Timeout and cleanup worker for background maintenance tasks.
This worker runs scheduled maintenance tasks:
- Approval timeout monitoring (every 30 seconds)
- S3 temp file cleanup (every hour)
- Orphaned job cleanup (every 4 hours)
- Metrics cleanup (daily)
"""
import asyncio
import logging
from datetime import UTC, datetime
from ..config import settings
from ..services.cleanup_service import CleanupService
from ..services.job_service import JobService
from ..services.metrics_service import (
MetricsService,
worker_active_gauge,
worker_errors_total,
)
from ..services.orphan_service import OrphanService
from ..services.queue_service import QueueService
from ..services.s3_cleanup_service import S3CleanupService
from ..services.storage_service import StorageService
from ..services.timeout_service import TimeoutService
logger = logging.getLogger(__name__)
class TimeoutWorker:
"""Background worker for timeout and cleanup operations."""
def __init__(
self,
storage_service: StorageService,
queue_service: QueueService,
job_service: JobService,
metrics_service: MetricsService,
s3_cleanup_service: S3CleanupService,
) -> None:
"""Initialize timeout worker with service dependencies.
Args:
storage_service: S3 storage operations
queue_service: Redis queue operations
job_service: Job status management
metrics_service: Metrics tracking
s3_cleanup_service: S3 cleanup operations (best-effort)
"""
# Core services
self.storage_service = storage_service
self.queue_service = queue_service
self.job_service = job_service
self.metrics_service = metrics_service
self.s3_cleanup_service = s3_cleanup_service
# Derived services
self.cleanup_service = CleanupService(storage_service)
self.timeout_service = TimeoutService(queue_service, job_service, self.cleanup_service, metrics_service)
self.orphan_service = OrphanService(job_service, self.s3_cleanup_service, metrics_service)
# Task scheduling state
self.last_approval_check: datetime | None = None
self.last_temp_cleanup: datetime | None = None
self.last_debug_cleanup: datetime | None = None
self.last_orphan_cleanup: datetime | None = None
self.last_metrics_cleanup: datetime | None = None
self.last_scaling_metrics_publish: datetime | None = None
# Worker state
self.running = False
async def start(self, shutdown_event: asyncio.Event | None = None) -> None:
"""Main worker loop with scheduled task execution.
Args:
shutdown_event: Optional event to signal graceful shutdown
"""
self.running = True
logger.info("Timeout worker started")
# Mark worker as active in Prometheus
worker_active_gauge.labels(worker_name="timeout").set(1)
try:
while self.running and (shutdown_event is None or not shutdown_event.is_set()):
try:
current_time = datetime.now(UTC)
# Task 1: Approval timeout monitoring (every 30 seconds)
if self._should_run_task(self.last_approval_check, settings.approval_check_interval_seconds):
await self._run_approval_check()
self.last_approval_check = current_time
# Task 2: S3 temp file cleanup (every hour)
if self._should_run_task(self.last_temp_cleanup, settings.temp_cleanup_interval_hours * 3600):
await self._run_temp_cleanup()
self.last_temp_cleanup = current_time
# Task 3: Debug artifact cleanup (every hour, same as temp cleanup)
if self._should_run_task(self.last_debug_cleanup, settings.temp_cleanup_interval_hours * 3600):
await self._run_debug_cleanup()
self.last_debug_cleanup = current_time
# Task 4: Orphaned job cleanup (every 4 hours)
if self._should_run_task(self.last_orphan_cleanup, settings.orphan_cleanup_interval_hours * 3600):
await self._run_orphan_cleanup()
self.last_orphan_cleanup = current_time
# Task 5: Metrics cleanup (daily)
if self._should_run_task(self.last_metrics_cleanup, settings.metrics_cleanup_interval_hours * 3600):
await self._run_metrics_cleanup()
self.last_metrics_cleanup = current_time
# Task 6: Publish scaling metrics to CloudWatch (every 60s, production only)
if settings.environment == "production" and self._should_run_task(
self.last_scaling_metrics_publish, 60
):
await self._publish_scaling_metrics()
self.last_scaling_metrics_publish = current_time
# Sleep before next iteration
await asyncio.sleep(settings.timeout_worker_check_interval_seconds)
except Exception as e:
logger.error(f"Error in timeout worker loop: {e}", exc_info=True)
# Track error
worker_errors_total.labels(worker_name="timeout", error_type=type(e).__name__).inc()
# Sleep longer on error to avoid rapid error loops
await asyncio.sleep(settings.timeout_worker_error_sleep_seconds)
except asyncio.CancelledError:
logger.info("Timeout worker received cancellation signal")
raise
finally:
# Mark worker as inactive when shutting down
worker_active_gauge.labels(worker_name="timeout").set(0)
logger.info("Timeout worker shutting down gracefully")
def _should_run_task(self, last_run: datetime | None, interval_seconds: int) -> bool:
"""Check if enough time has elapsed to run a scheduled task.
Args:
last_run: Datetime of last task execution (None if never run)
interval_seconds: Minimum seconds between task runs
Returns:
True if task should run
"""
if last_run is None:
return True
current_time = datetime.now(UTC)
elapsed = (current_time - last_run).total_seconds()
return elapsed >= interval_seconds
async def _run_approval_check(self) -> None:
"""Run approval timeout monitoring task."""
try:
logger.debug("Running approval timeout check...")
processed_count = await self.timeout_service.process_expired_approvals()
if processed_count > 0:
logger.info(f"Processed {processed_count} expired approvals")
except Exception as e:
logger.error(f"Error in approval timeout check: {e}", exc_info=True)
await self.metrics_service.increment_metric("worker_task_errors", 1)
async def _run_temp_cleanup(self) -> None:
"""Run S3 temp file cleanup task."""
try:
logger.info("Running temp file cleanup...")
result = await self.s3_cleanup_service.cleanup_expired_temp_files()
logger.info(
f"Temp cleanup complete: {result['files_deleted']} files deleted, "
f"{result['bytes_freed']} bytes freed, {result['errors']} errors"
)
except Exception as e:
logger.error(f"Error in temp file cleanup: {e}", exc_info=True)
await self.metrics_service.increment_metric("worker_task_errors", 1)
async def _run_debug_cleanup(self) -> None:
"""Run debug artifact cleanup task."""
try:
logger.info("Running debug artifact cleanup...")
result = await self.s3_cleanup_service.cleanup_expired_debug_artifacts()
logger.info(
f"Debug cleanup complete: {result['files_deleted']} files deleted, "
f"{result['bytes_freed']} bytes freed, {result['errors']} errors"
)
except Exception as e:
logger.error(f"Error in debug artifact cleanup: {e}", exc_info=True)
await self.metrics_service.increment_metric("worker_task_errors", 1)
async def _run_orphan_cleanup(self) -> None:
"""Run orphaned job cleanup tasks."""
try:
logger.info("Running orphan cleanup...")
# Task 1: Cleanup old completed/failed jobs
old_jobs_result = await self.orphan_service.cleanup_old_completed_jobs()
logger.info(
f"Old jobs cleanup: {old_jobs_result['jobs_cleaned']} cleaned, {old_jobs_result['errors']} errors"
)
# Task 2: Detect and fail stuck processing jobs
stuck_jobs_result = await self.orphan_service.cleanup_stuck_processing_jobs()
logger.info(
f"Stuck jobs detection: {stuck_jobs_result['jobs_failed']} failed, {stuck_jobs_result['errors']} errors"
)
except Exception as e:
logger.error(f"Error in orphan cleanup: {e}", exc_info=True)
await self.metrics_service.increment_metric("worker_task_errors", 1)
async def _run_metrics_cleanup(self) -> None:
"""Run metrics cleanup task."""
try:
logger.info("Running metrics cleanup...")
# Cleanup old metrics
deleted_count = await self.metrics_service.cleanup_old_metrics()
logger.info(f"Metrics cleanup complete: {deleted_count} keys deleted")
# Log daily summary (useful for monitoring)
await self.metrics_service.log_daily_summary()
except Exception as e:
logger.error(f"Error in metrics cleanup: {e}", exc_info=True)
await self.metrics_service.increment_metric("worker_task_errors", 1)
async def _publish_scaling_metrics(self) -> None:
"""Publish jobs-in-processing count to CloudWatch for docling auto-scaling."""
try:
redis = self.job_service.redis
raw = await redis.get("eq-pdf:metrics:jobs_in_processing")
count = max(0, int(raw or 0))
import boto3
cw = boto3.client("cloudwatch", region_name=settings.aws_region)
cw.put_metric_data(
Namespace="EqualifyPDF",
MetricData=[{
"MetricName": "JobsInProcessing",
"Value": count,
"Unit": "Count",
}],
)
logger.debug("Published scaling metric: JobsInProcessing=%d", count)
except Exception as e:
logger.warning("Failed to publish scaling metrics: %s", e)
def stop(self) -> None:
"""Signal worker to stop gracefully."""
logger.info("Stopping timeout worker...")
self.running = False
# Global worker instance
_worker_instance: TimeoutWorker | None = None
async def start_timeout_worker(shutdown_event: asyncio.Event | None = None) -> None:
"""Entry point for timeout worker (called from main.py lifespan).
Creates service instances and starts the worker loop.
This function is called from FastAPI lifespan context.
Args:
shutdown_event: Optional event to signal graceful shutdown
"""
global _worker_instance
logger.info("Initializing timeout worker...")
# Import dependencies dynamically to avoid circular imports
from ..dependencies import get_redis_client, get_s3_client
from ..services.job_service import JobService
from ..services.queue_service import QueueService
from ..services.storage_service import StorageService
# Get Redis and S3 clients using proper async generator pattern
redis_client = await anext(get_redis_client())
s3_client = await anext(get_s3_client())
# Create service instances directly with clients
storage_service = StorageService(
s3_client=s3_client,
temp_bucket=settings.s3_temp_bucket,
results_bucket=settings.s3_results_bucket,
)
queue_service = QueueService(redis_client=redis_client)
job_service = JobService(redis_client=redis_client)
# Metrics service uses the same Redis client
metrics_service = MetricsService(redis_client)
# Create S3 cleanup service (best-effort, no circuit breakers)
s3_cleanup_service = S3CleanupService(
s3_client=s3_client,
temp_bucket=settings.s3_temp_bucket,
)
# Create worker
_worker_instance = TimeoutWorker(
storage_service=storage_service,
queue_service=queue_service,
job_service=job_service,
metrics_service=metrics_service,
s3_cleanup_service=s3_cleanup_service,
)
logger.info("Timeout worker services initialized, starting worker loop...")
# Start worker (runs until stopped)
await _worker_instance.start(shutdown_event)
def get_timeout_worker() -> TimeoutWorker | None:
"""Get the global timeout worker instance.
Returns:
Optional[TimeoutWorker]: Worker instance if started, None otherwise
"""
return _worker_instance