move_detect/edge/alert.py
유창욱 2ea0420de7 fix(delivery): resolve codex + workflow review findings across edge/cloud/deploy
Two independent adversarial reviews (codex CLI + a 6-dimension verification
workflow) of the delivery-hardening branch surfaced ~40 findings; this closes
the confirmed correctness/security/deploy blockers plus a codex re-review pass.

edge:
- geometry: track_displacement per-second threshold recalibrated (0.02->0.25,
  env-tunable via MOTION_THR, validated finite/positive); confidence-filter
  keypoints with a bbox-center fallback so hallucinated points don't fake motion
- fall: low-confidence frames no longer cancel an active FALLING/FALLEN incident
  (freeze + reset recovery timer so recovery needs a fresh confident streak);
  baseline height only seeded from trusted frames
- queue: PostClient.send() returns a tri-state SendResult; only definite 4xx
  count toward the poison drop cap, so a network outage (incl. 408/425/429/5xx)
  never deletes queued safety events; legacy rows persist their minted uuid
- config: reject placeholder / http / short DEVICE_SECRET at startup
- zones: warn+skip legacy pixel-space polygons that would silently match nothing

cloud:
- zone endpoints wired into P3-2b site-scope (cross-site read/write/delete
  closed) + role gate on list; catch narrowed so DB faults surface as 500
- snapshot purge keeps the DB row when unlink() fails (no untrackable PIPA orphan)
- ingest dedup path re-stores missing snapshots (+ short-write guard); UNIQUE
  (event_id, content_hash) makes recovery idempotent under retry/race
- event list cursor carries full-precision occurred_at (no float rounding at the
  page boundary); string keyset end-to-end
- SiteScope '' fails closed (only NULL is global); dashboard fetches full page cap
- runnable migration (site/event_uuid + backfill ordering + indexes); rate-limit
  reap fails open

deploy:
- offline bundle hashes EVERY file (fatal on missing/failed/injected; models in
  SBOM; arch in tarball name; REQUIRE_SIGNATURE option)
- installer verifies before any mutation, Hailo arch preflight before install,
  atomic src swap, backend reconcile preserves a working operator choice
- systemd DeviceAllow=char-dma_heap (the /dev/dma_heap/ dir form was inert)

Verified: edge pytest 165 passed / 2 skipped; cloud phpunit 109 passed / 4
skipped; deploy scripts bash -n clean. Reviewed by codex twice (findings +
fixes) with the second-pass findings folded in.
2026-07-17 19:24:42 +09:00

79 lines
3.4 KiB
Python

import base64
import hashlib
import hmac
import json
from enum import Enum
from edge.events import Event
# Shared-hosting WAFs (cafe24) frequently block the default "python-requests/x.y" User-Agent
# with an HTML 403 while curl passes — so every event would silently 403 with a correct HMAC
# and URL. Send an explicit UA on every request. CloudClient imports this too.
USER_AGENT = "move-detect-edge/1.0"
class SendResult(Enum):
"""Outcome of one upload attempt. The offline queue must tell a transient outage (retry
forever — a real fall alert has to survive an overnight internet outage) apart from a
permanent rejection (poison — count toward the drop cap), which a bare bool cannot express.
DELIVERED is the only truthy value, so `if client.send(...)` keeps working unchanged."""
DELIVERED = "delivered"
REJECTED = "rejected" # a definite 4xx (not 429): the server will never accept this event
TRANSIENT = "transient" # network error / timeout / 429 / 5xx — connectivity or server hiccup
def __bool__(self) -> bool:
return self is SendResult.DELIVERED
def sign(secret: str, nonce: str, ts: str, body: str | bytes) -> str:
if isinstance(body, str):
body = body.encode()
msg = nonce.encode() + ts.encode() + body
return hmac.new(secret.encode(), msg, hashlib.sha256).hexdigest()
class PostClient:
def __init__(self, url: str, device_id: str, secret: str,
nonce_fn, now_fn, http_post=None, timeout: float = 5.0):
self.url = url
self.device_id = device_id
self.secret = secret
self.nonce_fn = nonce_fn
self.now_fn = now_fn
self.timeout = timeout
if http_post is None:
import requests
http_post = requests.post
self._post = http_post
def send(self, event: Event, snapshots: list[bytes]) -> SendResult:
payload = event.to_payload(self.device_id)
payload["snapshots"] = [base64.b64encode(s).decode() for s in snapshots]
body = json.dumps(payload, sort_keys=True)
nonce = self.nonce_fn()
ts = str(self.now_fn())
headers = {
"Content-Type": "application/json",
"User-Agent": USER_AGENT,
"X-Nonce": nonce,
"X-Timestamp": ts,
"X-Signature": sign(self.secret, nonce, ts, body.encode()),
}
try:
# allow_redirects=False: the .htaccess http->https 302 must not silently become a
# GET (405/404). A redirect is a CAFE24_URL misconfiguration to surface, not follow.
resp = self._post(self.url, data=body, headers=headers, timeout=self.timeout,
allow_redirects=False)
except Exception:
# Connection refused / DNS / timeout — the cloud is unreachable, not rejecting us.
return SendResult.TRANSIENT
code = resp.status_code
if 200 <= code < 300:
return SendResult.DELIVERED
# Transient — "back off and retry", NOT a permanent verdict on the event: all 5xx plus the
# 4xx codes that explicitly mean try-again (408 Request Timeout, 425 Too Early, 429 Too
# Many Requests). Counting these toward the poison drop cap would delete real events during
# a proxy/server hiccup. Only a definite "bad request" 4xx is a permanent rejection.
if code in (408, 425, 429) or code >= 500:
return SendResult.TRANSIENT
return SendResult.REJECTED