The analysis-disabled gate (stop/inhibit vision when Admin→Settings analysis
is OFF, so stale queued jobs don't pin ~4.4GiB) was applied directly on the
host and never committed, so it was absent from main and from this branch —
merging as-is would have regressed it. Re-introduce it here so the merged
manager has BOTH gates: vision is held down if analysis is disabled OR host
memory is tight. Also grant the read-only role SELECT on app_settings.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a fourth gate to the vision autoscaler poll loop: when host memory
gets tight, proactively SIGTERM-stop the mangalord-vision container so a
transient spike elsewhere (a crawl, CI, a backend burst) can finish
instead of the kernel OOM-killer shooting something stateful.
- read_mem_used_pct: host-wide used% from /proc/meminfo MemAvailable
(not MemFree), ERR sentinel when MemTotal or MemAvailable is missing.
- mem_check_and_stop: active stop over MEM_HIGH_WATERMARK_PCT + arm a
restart cooldown; only inspects docker once already over the mark.
- mem_yield_active: start inhibit while cooldown active or used% >=
MEM_LOW_WATERMARK_PCT (two watermarks = hysteresis).
- Inhibit reuses the existing idle path via pending=0; flip-only logging;
stop_vision gains a reason arg so idle/MAX_UPTIME/memory-yield stops are
distinctly logged.
- Faster MEM_POLL_INTERVAL sub-poll between backlog ticks to catch spikes.
Backend verifications V1-V4 (see VISION-MEMORY-YIELD.md) all pass against
the current queue, so no backend changes are needed. RESPECT_CRAWL_MUTEX
is kept for now. New VISION_MEM_* knobs wired through docker-compose.yml
and documented in .env.example.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
