Spaces:
Running
Running
File size: 13,207 Bytes
a55ea57 2f1e7ff a55ea57 2f1e7ff a55ea57 2f1e7ff a55ea57 2f1e7ff a55ea57 2f1e7ff a55ea57 2f1e7ff a55ea57 2f1e7ff a55ea57 2f1e7ff a55ea57 2f1e7ff a55ea57 | 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 | import asyncio
import inspect
import logging
import os
import time
import traceback
from contextlib import contextmanager
from contextvars import ContextVar
from typing import Any, Iterator, Optional
from src.services.redaction import redact
from src.lib.tracing.events import TraceEvent, TraceEventType, TraceSeverity
from src.lib.tracing.subscribers import TraceSubscriber
logger = logging.getLogger(__name__)
_trace_id_var: ContextVar[Optional[str]] = ContextVar("trace_id", default=None)
_tenant_id_var: ContextVar[Optional[str]] = ContextVar("tenant_id", default=None)
class LocalTraceQueue:
def __init__(self, max_size: int = 10000):
if max_size <= 0:
raise ValueError("Trace queue capacity must be positive and bounded")
self._queue: asyncio.Queue[TraceEvent] = asyncio.Queue(maxsize=max_size)
@property
def size(self) -> int:
return self._queue.qsize()
def put_nowait(self, event: TraceEvent) -> bool:
try:
self._queue.put_nowait(event)
return True
except asyncio.QueueFull:
return False
async def get(self) -> TraceEvent:
return await self._queue.get()
def task_done(self) -> None:
self._queue.task_done()
def drain(self, limit: Optional[int] = None) -> list[TraceEvent]:
events: list[TraceEvent] = []
while not self._queue.empty() and (limit is None or len(events) < limit):
events.append(self._queue.get_nowait())
self._queue.task_done()
return events
class TraceTracker:
def __init__(
self,
*,
queue: Optional[LocalTraceQueue] = None,
max_queue_size: Optional[int] = None,
slow_operation_ms: Optional[float] = None,
downtime_ms: Optional[float] = None,
expensive_loop_iterations: Optional[int] = None,
enabled: Optional[bool] = None,
mode: Optional[str] = None,
):
self.mode = mode or ("queue" if queue is not None else os.getenv("TRACE_MODE", "log"))
if self.mode not in {"disabled", "log", "queue"}:
raise ValueError("TRACE_MODE must be disabled, log, or queue")
self.dropped_events = 0
self._last_drop_warning = float("-inf")
self.queue = queue or LocalTraceQueue(
max_queue_size
if max_queue_size is not None
else int(os.getenv("TRACE_QUEUE_MAX_SIZE", "10000"))
)
self.slow_operation_ms = (
slow_operation_ms
if slow_operation_ms is not None
else float(os.getenv("TRACE_SLOW_OPERATION_MS", "1000"))
)
self.downtime_ms = (
downtime_ms
if downtime_ms is not None
else float(os.getenv("TRACE_DOWNTIME_MS", "5000"))
)
self.expensive_loop_iterations = (
expensive_loop_iterations
if expensive_loop_iterations is not None
else int(os.getenv("TRACE_EXPENSIVE_LOOP_ITERATIONS", "10000"))
)
self.enabled = (
enabled
if enabled is not None
else os.getenv("TRACE_TRACKER_ENABLED", "true") == "true"
)
self._subscribers: list[TraceSubscriber] = []
self._subscriber_tasks: list[asyncio.Task] = []
@property
def default_sla_ms(self) -> float:
return self.slow_operation_ms
def set_context(
self, *, trace_id: Optional[str] = None, tenant_id: Optional[str] = None
):
trace_token = _trace_id_var.set(trace_id) if trace_id is not None else None
tenant_token = _tenant_id_var.set(tenant_id) if tenant_id is not None else None
return trace_token, tenant_token
def reset_context(self, tokens) -> None:
trace_token, tenant_token = tokens
if trace_token is not None:
_trace_id_var.reset(trace_token)
if tenant_token is not None:
_tenant_id_var.reset(tenant_token)
def subscribe(self, subscriber: TraceSubscriber) -> None:
self._subscribers.append(subscriber)
def clear_subscribers(self) -> None:
self._subscribers.clear()
def publish(
self,
event_type: TraceEventType | str,
name: str,
*,
severity: TraceSeverity | str = TraceSeverity.INFO,
tenant_id: Optional[str] = None,
trace_id: Optional[str] = None,
service: Optional[str] = None,
operation: Optional[str] = None,
duration_ms: Optional[float] = None,
threshold_ms: Optional[float] = None,
status_code: Optional[int] = None,
message: Optional[str] = None,
metadata: Optional[dict[str, Any]] = None,
exception: Optional[BaseException] = None,
) -> Optional[TraceEvent]:
if not self.enabled or self.mode == "disabled":
return None
event_trace_id = trace_id or _trace_id_var.get()
event_message = message or (str(exception) if exception else None)
event = TraceEvent(
event_type=TraceEventType(event_type),
severity=TraceSeverity(severity),
name=name,
service=service,
operation=operation,
tenant_id=tenant_id or _tenant_id_var.get(),
trace_id=event_trace_id,
duration_ms=duration_ms,
threshold_ms=threshold_ms,
status_code=status_code,
error_type=type(exception).__name__ if exception else None,
message=redact(event_message),
metadata=redact(dict(metadata or {})),
stack_trace=(
redact("".join(traceback.format_exception(exception))) if exception else None
),
)
if self.mode == "log":
level = {"info": logging.INFO, "warning": logging.WARNING, "error": logging.ERROR, "critical": logging.CRITICAL}[event.severity.value]
logger.log(level, "%s %s", event.name, event.to_dict())
return event
if not self.queue.put_nowait(event):
self.dropped_events += 1
now = time.monotonic()
if now - self._last_drop_warning >= 60:
logger.warning("Trace queue saturated; dropped events=%d", self.dropped_events)
self._last_drop_warning = now
return None
return event
def exception(
self,
name: str,
exception: BaseException,
*,
service: Optional[str] = None,
operation: Optional[str] = None,
tenant_id: Optional[str] = None,
trace_id: Optional[str] = None,
duration_ms: Optional[float] = None,
threshold_ms: Optional[float] = None,
status_code: Optional[int] = None,
metadata: Optional[dict[str, Any]] = None,
) -> Optional[TraceEvent]:
return self.publish(
TraceEventType.EXCEPTION,
name,
severity=TraceSeverity.ERROR,
service=service,
operation=operation,
tenant_id=tenant_id,
trace_id=trace_id,
duration_ms=duration_ms,
threshold_ms=threshold_ms,
status_code=status_code,
metadata=metadata,
exception=exception,
)
def error(
self,
name: str,
*,
service: Optional[str] = None,
operation: Optional[str] = None,
tenant_id: Optional[str] = None,
trace_id: Optional[str] = None,
severity: TraceSeverity | str = TraceSeverity.ERROR,
duration_ms: Optional[float] = None,
threshold_ms: Optional[float] = None,
status_code: Optional[int] = None,
message: Optional[str] = None,
metadata: Optional[dict[str, Any]] = None,
) -> Optional[TraceEvent]:
return self.publish(
TraceEventType.ERROR,
name,
severity=severity,
service=service,
operation=operation,
tenant_id=tenant_id,
trace_id=trace_id,
duration_ms=duration_ms,
threshold_ms=threshold_ms,
status_code=status_code,
message=message,
metadata=metadata,
)
def downtime(
self,
name: str,
duration_ms: float,
*,
service: Optional[str] = None,
operation: Optional[str] = None,
tenant_id: Optional[str] = None,
trace_id: Optional[str] = None,
metadata: Optional[dict[str, Any]] = None,
) -> Optional[TraceEvent]:
return self.publish(
TraceEventType.DOWNTIME,
name,
severity=TraceSeverity.CRITICAL,
service=service,
operation=operation,
tenant_id=tenant_id,
trace_id=trace_id,
duration_ms=duration_ms,
metadata=metadata,
)
def expensive_loop(
self,
name: str,
iterations: int,
*,
service: Optional[str] = None,
operation: Optional[str] = None,
tenant_id: Optional[str] = None,
trace_id: Optional[str] = None,
metadata: Optional[dict[str, Any]] = None,
threshold: Optional[int] = None,
) -> Optional[TraceEvent]:
limit = (
threshold
if threshold is not None
else self.expensive_loop_iterations
)
if iterations < limit:
return None
return self.publish(
TraceEventType.EXPENSIVE_LOOP,
name,
severity=TraceSeverity.WARNING,
service=service,
operation=operation,
tenant_id=tenant_id,
trace_id=trace_id,
metadata={**(metadata or {}), "iterations": iterations},
)
@contextmanager
def span(
self,
name: str,
*,
tenant_id: Optional[str] = None,
trace_id: Optional[str] = None,
metadata: Optional[dict[str, Any]] = None,
slow_operation_ms: Optional[float] = None,
service: Optional[str] = None,
operation: Optional[str] = None,
) -> Iterator[None]:
tokens = self.set_context(trace_id=trace_id, tenant_id=tenant_id)
started_at = time.perf_counter()
try:
yield
except Exception as exc:
duration_ms = (time.perf_counter() - started_at) * 1000
self.exception(
name,
exc,
service=service,
operation=operation,
metadata={**(metadata or {}), "duration_ms": duration_ms},
)
raise
finally:
duration_ms = (time.perf_counter() - started_at) * 1000
threshold = (
slow_operation_ms
if slow_operation_ms is not None
else self.slow_operation_ms
)
if duration_ms >= threshold:
self.publish(
TraceEventType.SLA_BREACH,
name,
severity=TraceSeverity.WARNING,
service=service,
operation=operation,
duration_ms=duration_ms,
threshold_ms=threshold,
metadata=metadata,
)
self.reset_context(tokens)
def track_loop(
self,
name: str,
iterable,
*,
tenant_id: Optional[str] = None,
trace_id: Optional[str] = None,
metadata: Optional[dict[str, Any]] = None,
threshold: Optional[int] = None,
):
count = 0
for item in iterable:
count += 1
yield item
if count >= (threshold or self.expensive_loop_iterations):
self.expensive_loop(
name,
count,
tenant_id=tenant_id,
trace_id=trace_id,
metadata=metadata,
threshold=threshold,
)
async def dispatch_once(self) -> TraceEvent:
event = await self.queue.get()
try:
for subscriber in list(self._subscribers):
result = subscriber.handle(event)
if inspect.isawaitable(result):
await result
finally:
self.queue.task_done()
return event
async def subscribe_forever(self) -> None:
while True:
await self.dispatch_once()
def start_subscribers(self) -> None:
if not self._subscribers or self._subscriber_tasks:
return
try:
loop = asyncio.get_running_loop()
except RuntimeError:
return
self._subscriber_tasks = [
loop.create_task(self.subscribe_forever()) for _ in self._subscribers
]
async def stop_subscribers(self) -> None:
for task in self._subscriber_tasks:
task.cancel()
for task in self._subscriber_tasks:
try:
await task
except asyncio.CancelledError:
pass
self._subscriber_tasks = []
trace_tracker = TraceTracker()
tracer = trace_tracker
|