bvandeusenandClaude Opus 5.5
b8f73270e7
The dashboard's host widgets took 18-60 s. Every "latest reading" query (hosts overview, host-agent fleet, host page, vitals, SNMP, MCP get_host) searched plugin_metrics for the newest row per metric. Postgres has no skip scan, so each one read every raw sample in the 7-day window. - plugin_metric_latest: one row per (source, resource, metric). record_metrics upserts it in the same transaction, moving forward only and in sorted key order. Migration 0030 seeds it by stepping the existing index. - Retention drops a latest row once its sample leaves the raw window, so a resource that stopped reporting disappears when its history does, as before. Renames carry it along. - A quiet host keeps its last values, marked stale. A time-bounded query would have shown it as missing. - Vitals sparklines span the 6 h they claim, averaged in SQL to ~120 points, instead of fetching 6 h raw and drawing the last 120 samples. - Dashboard and share-view cells: "Loading…" until the first answer; a 20 s deadline per fetch; "Couldn't load (why)" with Retry now and an automatic retry for cells that don't poll; "as of HH:MM" in the header, and "couldn't refresh · as of" when a refresh fails over shown data. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
85 lines
3.8 KiB
Python
85 lines
3.8 KiB
Python
"""steward/core/renames.py
|
|
|
|
Alert rules, maintenance-window scopes and metric history all key a thing
|
|
by (source, name). When a monitor or host is renamed, carry_rename moves
|
|
every one of them to the new name, in the caller's transaction, so the
|
|
thing's rules keep watching it and its history stays in one piece (#5065).
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
from sqlalchemy import and_, delete, func, or_, update
|
|
from sqlalchemy.ext.asyncio import AsyncSession
|
|
|
|
from steward.models.alerts import AlertRule, MaintenanceWindow
|
|
from steward.models.metrics import PluginMetric, PluginMetricHourly, PluginMetricLatest
|
|
|
|
# What a host's name keys, and how each source names a part of the host:
|
|
# host_agent "web01:/var", "web01:core3"; docker "web01/nginx".
|
|
HOST_SOURCES: tuple[tuple[str, str | None], ...] = (("host_agent", ":"), ("docker", "/"))
|
|
|
|
|
|
def _matches(source_col, name_col, sources, name: str):
|
|
"""(source, name) itself, or a sub-resource "<name><sep>…" of it."""
|
|
clauses = []
|
|
for source, sep in sources:
|
|
under = [name_col == name]
|
|
if sep:
|
|
# An exact prefix: LIKE would read % or _ in a name as wildcards, and
|
|
# a >=/< range follows the collation, which may ignore punctuation.
|
|
under.append(func.left(name_col, len(name) + len(sep)) == name + sep)
|
|
clauses.append(and_(source_col == source, or_(*under)))
|
|
return or_(*clauses)
|
|
|
|
|
|
def _renamed(name_col, old: str, new: str):
|
|
return func.concat(new, func.substr(name_col, len(old) + 1))
|
|
|
|
|
|
async def carry_rename(session: AsyncSession, sources, old: str, new: str) -> dict[str, int]:
|
|
"""Point everything keyed by `old` under each (source, separator) in
|
|
`sources` at `new`. A separator also carries the sub-resources named
|
|
"<old><sep>…"; None carries the exact name only.
|
|
|
|
The caller has already refused a `new` that a live thing holds, so any
|
|
history under `new` belongs to something deleted: it is dropped first,
|
|
rather than merged into this thing's history. Returns rows moved, by kind.
|
|
"""
|
|
if not old or old == new:
|
|
return {}
|
|
sources = tuple(sources)
|
|
for model in (PluginMetric, PluginMetricHourly, PluginMetricLatest):
|
|
await session.execute(delete(model).where(
|
|
_matches(model.source_module, model.resource_name, sources, new)))
|
|
moved = {}
|
|
for kind, model, src, res in (
|
|
("rules", AlertRule, AlertRule.source_module, AlertRule.resource_name),
|
|
("maintenance_windows", MaintenanceWindow, MaintenanceWindow.scope_source,
|
|
MaintenanceWindow.scope_resource),
|
|
("metric_rows", PluginMetric, PluginMetric.source_module, PluginMetric.resource_name),
|
|
("hourly_rows", PluginMetricHourly, PluginMetricHourly.source_module,
|
|
PluginMetricHourly.resource_name),
|
|
("latest_rows", PluginMetricLatest, PluginMetricLatest.source_module,
|
|
PluginMetricLatest.resource_name),
|
|
):
|
|
result = await session.execute(
|
|
update(model).where(_matches(src, res, sources, old))
|
|
.values({res: _renamed(res, old, new)})
|
|
.execution_options(synchronize_session=False))
|
|
moved[kind] = result.rowcount
|
|
return moved
|
|
|
|
|
|
def describe(moved: dict[str, int]) -> str:
|
|
"""'2 alert rules, 1 maintenance window and 18,204 metric samples', or ''."""
|
|
parts = []
|
|
for kind, one, many in (("rules", "alert rule", "alert rules"),
|
|
("maintenance_windows", "maintenance window", "maintenance windows")):
|
|
if moved.get(kind):
|
|
parts.append(f"{moved[kind]} {one if moved[kind] == 1 else many}")
|
|
samples = moved.get("metric_rows", 0) + moved.get("hourly_rows", 0)
|
|
if samples:
|
|
parts.append(f"{samples:,} metric sample{'' if samples == 1 else 's'}")
|
|
if len(parts) > 1:
|
|
return ", ".join(parts[:-1]) + " and " + parts[-1]
|
|
return "".join(parts)
|