perf(monitors): a monitor's latest results by a LATERAL probe, not row_number over its history (#5208)
CI / lint (push) Successful in 3s
CI / integration (push) Successful in 49s
CI / unit (push) Successful in 1m17s
CI / publish (push) Successful in 1m30s

_last_n_results (Status, the monitors list, host monitors), host_status
(hosts overview, list_hosts) and the report's down-monitor list each
ranked every result a monitor ever recorded to keep the newest few. Each
monitor now gets ORDER BY checked_at DESC LIMIT n in a LATERAL subquery,
which walks ix_monitor_results_monitor_checked backwards and stops.

Also fixes the latest-table test, which read between two begin() blocks
and so autobegan a transaction the second block then refused.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
2026-10-06 23:08:39 -04:00
co-authored by Claude Opus 5.5
parent b8f73270e7
commit 1845ba973c
5 changed files with 77 additions and 50 deletions
+2 -1
View File
@@ -54,7 +54,8 @@ def test_the_newest_sample_wins_and_a_late_backlog_does_not_move_it_back(app):
_sample(host, "cpu_pct", 20.0, now - timedelta(minutes=1)),
_sample(host, "mem_used_pct", 40.0, now - timedelta(minutes=1)),
])
after_first = await _latest(s, host)
async with s.begin():
after_first = await _latest(s, host)
async with s.begin():
# An agent catching up posts older samples after newer ones.
await record_metrics(s, [
+42
View File
@@ -83,3 +83,45 @@ def test_status_source_rolls_up_results(app):
assert e.status == "up" # latest result is up
assert len(e.heartbeat) == 4
assert e.uptime["24h"] == 75.0 # 3 of 4 up
@_NEEDS_DB
def test_last_results_are_the_newest_n_oldest_first_and_host_status_reads_the_last(app):
"""_last_n_results takes each monitor's newest n by a LATERAL probe (#5208);
host_status reads the newest one per enabled monitor."""
from steward.core.status import _last_n_results
from steward.hosts.service import host_status
from steward.models.hosts import Host
from steward.models.monitors import Monitor, MonitorResult
now = datetime.now(timezone.utc)
hid, a, b, quiet = (str(uuid.uuid4()) for _ in range(4))
async def _go():
async with app.db_sessionmaker() as s:
async with s.begin():
s.add(Host(id=hid, name="lateral-" + hid[:8], address="10.8.8.8"))
async with s.begin():
s.add(Monitor(id=a, name="a-" + a[:8], type="icmp", target="10.8.8.8",
host_id=hid, config_json="{}", enabled=True))
s.add(Monitor(id=b, name="b-" + b[:8], type="icmp", target="10.8.8.9",
host_id=hid, config_json="{}", enabled=True))
s.add(Monitor(id=quiet, name="q-" + quiet[:8], type="icmp", target="10.8.8.7",
config_json="{}", enabled=True))
async with s.begin():
for i in range(5): # a: up, up, up, up, down (newest)
s.add(MonitorResult(monitor_id=a, checked_at=now - timedelta(minutes=5 - i),
is_up=i < 4))
s.add(MonitorResult(monitor_id=b, checked_at=now - timedelta(minutes=1),
is_up=True, response_ms=12.0))
async with s.begin():
last = await _last_n_results(s, [a, b, quiet], n=3)
status = await host_status(s, [hid])
return last, status
last, status = asyncio.run(_go())
assert [r.is_up for r in last[a]] == [True, True, False]
assert last[a][0].checked_at < last[a][-1].checked_at # oldest first
assert len(last[b]) == 1 and last[quiet] == []
assert status[hid]["state"] == "down" and status[hid]["count"] == 2
assert status[hid]["latency_ms"] == 12.0