feat(alerts): an acknowledgement is on the record, with who made it (#5463)
CI / lint (push) Successful in 3s
CI / integration (push) Successful in 1m7s
CI / unit (push) Successful in 1m27s
CI / publish (push) Successful in 1m8s

#5463 "Acknowledgements on record: an event and who did it", step 1 of
milestone 499 "Alerts you can trust". The live walk in #5461 found that
acknowledging changed only the alert's current state: once it resolved,
nothing said who had acknowledged it or when.

- alert_events.actor_id (migration 0040): the person behind a transition.
- Acknowledging writes a firing → acknowledged event by the acknowledger.
- A rule edit, disable or enable that resets an alert carries its actor
  (web and MCP alike).
- The history page notes "acknowledged by X" and "rule disabled by X"; MCP
  get_alert_history returns `by`.
- One usernames() helper in alerts.service, used by the active list, the
  history page and the MCP tools.
- docs/core/alerting.md: the state diagram no longer shows an acknowledged
  alert re-firing, and says what an event records.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
2026-10-09 08:19:48 -04:00
co-authored by Claude Opus 5.5
parent b3a5bbfeff
commit 734d313a54
9 changed files with 133 additions and 48 deletions
+17 -6
View File
@@ -1,6 +1,7 @@
"""Integration: alert state against real Postgres. An acknowledgement holds
while the breach lasts, an alert clears after its OK run, and changing what a
rule watches (or disabling it) resets its state with a note on the event."""
while the breach lasts and is on the record with who made it (#5463), an alert
clears after its OK run, and changing what a rule watches (or disabling it)
resets its state with a note and the person on the event."""
from __future__ import annotations
import asyncio
@@ -44,7 +45,7 @@ def test_acknowledge_holds_clear_after_and_reset(app):
metric_name="cpu_pct", operator=">", threshold=90,
consecutive_failures_required=1, consecutive_ok_required=2)
rule = await alert_svc.add_rule(s, fields, created_by=user.id)
rule_id = rule.id
rule_id, user_id = rule.id, user.id
async def sample(v):
async with app.db_sessionmaker() as s:
@@ -62,11 +63,18 @@ def test_acknowledge_holds_clear_after_and_reset(app):
.order_by(AlertEvent.transitioned_at))).scalars().all()
return [(e.from_state, e.to_state, e.note) for e in rows]
async def last_actor():
async with app.db_sessionmaker() as s:
return (await s.execute(select(AlertEvent.actor_id).where(
AlertEvent.rule_id == rule_id).order_by(
AlertEvent.transitioned_at.desc()).limit(1))).scalar_one()
await sample(95)
assert (await state()).state == AlertStateEnum.firing
async with app.db_sessionmaker() as s:
async with s.begin():
await alert_svc.acknowledge(s, rule_id, None)
await alert_svc.acknowledge(s, rule_id, user_id)
assert await last_actor() == user_id
for _ in range(3):
await sample(97) # still breached: the acknowledgement holds
assert (await state()).state == AlertStateEnum.acknowledged
@@ -75,7 +83,9 @@ def test_acknowledge_holds_clear_after_and_reset(app):
await sample(10)
st = await state()
assert st.state == AlertStateEnum.inactive and st.acknowledged_at is None
assert await events() == [("inactive", "firing", None), ("acknowledged", "resolved", None)]
assert await events() == [("inactive", "firing", None), ("firing", "acknowledged", None),
("acknowledged", "resolved", None)]
assert await last_actor() is None # the readings resolved it, not a person
# Changing the threshold of a firing rule resets it, saying why.
await sample(95)
@@ -100,9 +110,10 @@ def test_acknowledge_holds_clear_after_and_reset(app):
async with app.db_sessionmaker() as s:
async with s.begin():
r = await s.get(AlertRule, rule_id)
await alert_svc.apply_rule_change(s, r, {}, enabled=False)
await alert_svc.apply_rule_change(s, r, {}, enabled=False, actor_id=user_id)
assert (await state()).state == AlertStateEnum.inactive
assert (await events())[-1] == ("firing", "resolved", "rule disabled")
assert await last_actor() == user_id
# Re-enabling starts the watch over: silence is measured from now.
async with app.db_sessionmaker() as s: