Erik
3d52e0b79e
press_settings.presses and settings.downtime_reasons are re-read on every 12s overview poll; route them through the kernel cached_select() TTL cache so they stop consuming a shared-pool checkout each poll. Part of the gateway connection-pool-saturation fix (see fastAPI/agent_docs/db-pool-relief.md). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
22 lines
985 B
Python
22 lines
985 B
Python
"""Kernel DB seam for the temper_overview page-module.
|
|
|
|
Mirrors ``modules/fastpress/press/_db.py``: the module reaches the database ONLY
|
|
through the gateway kernel's SELECT-only proxy, re-exported here so the coupling
|
|
to ``app.db`` lives in exactly one place. This page is read-only, so it needs
|
|
``run_select_query`` and nothing else (no ``transaction`` — it never writes).
|
|
|
|
The base scaffold's API (``api/routes.py``) returns SAMPLE data and does NOT import
|
|
this yet — wiring real SQL is the next step (see the TODO in ``api/routes.py``).
|
|
Keeping the seam here means that when the real query is added it imports::
|
|
|
|
from .._db import run_select_query
|
|
|
|
NOTE: no connection happens at import — SQLAlchemy's engine is lazy; the first
|
|
real connection opens only when a query actually executes.
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
from app.db.mariadb import run_select_query, cached_select # noqa: F401 (kernel read proxies)
|
|
|
|
__all__ = ["run_select_query", "cached_select"]
|