bvandeusenandClaude Opus 5.5
3c24c633df
- One run per host: start_run claims every host it touches (addresses read from the inventory Steward wrote; a source's own inventory file is claimed as that file) before it takes a run slot. The claim is all-or-nothing. A run waiting on a host shows queued and names the run it waits for, honours a cancel, and fails after HOST_WAIT_SECONDS (1 h) with which host and run. - A due schedule whose previous run is still queued or running skips the firing: last_run_at moves on, last_run_id still names the run going, and last_error says "Skipped: the previous run was still going". The Schedules page shows that beside the run (muted), and a launch error in red. - An alert rule that fires while the action run from an earlier firing is still going starts no new run; the event says so and names the run. An in-process set closes the gap between the check and the run being linked. - docs/core/ansible.md: queued/cancelled statuses and the overlap rules. Tests: tests/core/test_ansible_overlap.py (claim keys, waiting, all-or-nothing, cancel, deadline) and tests/integration/test_ansible_overlap.py (two real runs on one host take turns; schedule skip then fire; alert action skip then start). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
83 lines
3.0 KiB
Python
83 lines
3.0 KiB
Python
# steward/ansible/scheduler.py
|
|
"""Fire due Ansible schedules. Driven by the core ScheduledTask loop (~60s)."""
|
|
from __future__ import annotations
|
|
|
|
import logging
|
|
from datetime import datetime, timezone
|
|
|
|
from sqlalchemy import select
|
|
|
|
from steward.ansible.runner import run_in_flight, trigger_run
|
|
from steward.models.ansible_schedule import AnsibleSchedule
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
def is_due(schedule: AnsibleSchedule, now: datetime) -> bool:
|
|
"""True when an enabled schedule has never run or its interval has elapsed.
|
|
|
|
Cadence is measured from the previous *fire* time (last_run_at), not run
|
|
completion, so a long-running playbook doesn't cause catch-up storms.
|
|
"""
|
|
if not schedule.enabled:
|
|
return False
|
|
if schedule.last_run_at is None:
|
|
return True
|
|
return (now - schedule.last_run_at).total_seconds() >= schedule.interval_seconds
|
|
|
|
|
|
async def run_due_schedules(app) -> int:
|
|
"""Trigger every due schedule once. Returns the number launched. A due
|
|
schedule whose previous run hasn't finished skips this firing."""
|
|
now = datetime.now(timezone.utc)
|
|
async with app.db_sessionmaker() as db:
|
|
schedules = (await db.execute(
|
|
select(AnsibleSchedule).where(AnsibleSchedule.enabled.is_(True))
|
|
)).scalars().all()
|
|
|
|
fired = 0
|
|
for s in schedules:
|
|
if not is_due(s, now):
|
|
continue
|
|
async with app.db_sessionmaker() as db:
|
|
busy = await run_in_flight(db, s.last_run_id)
|
|
if busy:
|
|
# One run at a time per schedule (#5158). This firing is skipped,
|
|
# not deferred, so the cadence holds; last_run_id still names the
|
|
# run that's going.
|
|
async with app.db_sessionmaker() as db:
|
|
async with db.begin():
|
|
fresh = await db.get(AnsibleSchedule, s.id)
|
|
if fresh:
|
|
fresh.last_run_at = now
|
|
fresh.last_error = "Skipped: the previous run was still going"
|
|
logger.info("Schedule %s (%s) skipped: run %s still going",
|
|
s.id, s.name, s.last_run_id)
|
|
continue
|
|
run = None
|
|
err: str | None = None
|
|
try:
|
|
run, _source, err = await trigger_run(
|
|
app,
|
|
source_name=s.source_name,
|
|
playbook_path=s.playbook_path,
|
|
inventory_scope=s.inventory_scope,
|
|
params=s.params,
|
|
triggered_by=None,
|
|
)
|
|
except Exception:
|
|
logger.exception("Schedule %s (%s) failed to launch", s.id, s.name)
|
|
err = "launch error"
|
|
|
|
async with app.db_sessionmaker() as db:
|
|
async with db.begin():
|
|
fresh = await db.get(AnsibleSchedule, s.id)
|
|
if fresh:
|
|
fresh.last_run_at = now
|
|
fresh.last_run_id = run.id if run else None
|
|
fresh.last_error = err
|
|
if run:
|
|
fired += 1
|
|
logger.info("Schedule %s (%s) fired run %s", s.id, s.name, run.id)
|
|
return fired
|