OOM-killed every 16 hours — telemetry buffer grew to 2 GB

The code that's running in prod

# telemetry.py
# Background telemetry buffer — accumulates events from every
# request and flushes to an external ingestion endpoint every 60s.

import threading
import time
import requests

_buffer: list = []
_lock = threading.Lock()


def record_event(event):
    """Called from every request handler."""
    with _lock:
        _buffer.append(event)


def flush_loop():
    """Background thread: ship the buffer every 60 seconds."""
    while True:
        time.sleep(60)
        with _lock:
            if not _buffer:
                continue
            batch = list(_buffer)
        try:
            requests.post(
                "https://telemetry.example.com/ingest",
                json=batch,
                timeout=5,
            )
            with _lock:
                _buffer.clear()
        except Exception:
            pass   # try again on the next cycle

Log dashboard (last 60 seconds)

Pick the root cause

Four candidates. Three are plausible but mistake symptom for cause. Pick what you'd write in the postmortem timeline.