feat(dashboard): per-metric sparklines in the Host Agent — Resources widget
CI / lint (push) Successful in 3s
CI / unit (push) Successful in 9s
CI / integration (push) Successful in 2m17s
CI / publish (push) Successful in 47s

Operator ask: show the graphs next to their fields in the fleet widget, like the
host page's AGENT panel. Each row now renders cpu/mem/disk/load as a value with a
trend sparkline beneath it (1h window), instead of bare numbers.

- _fleet_rows fetches a per-host recent series (cpu/mem/load host-level, disk from
  the root mount) in one 1h query and attaches a sparkline per metric to each row.
- widget_table.html lays out a metric cell (label + threshold-coloured value +
  sparkline) per field, mirroring panel.html. Threshold colour is computed in the
  loop and passed into the cell macro (keeps Jinja macro scope clean).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-06-17 16:06:05 -04:00
parent 88dca32d3c
commit cb47b5e977
3 changed files with 59 additions and 18 deletions
+39
View File
@@ -349,6 +349,39 @@ async def _fleet_rows(session) -> list[dict]:
continue
latest.setdefault(row.resource_name, {})[row.metric_name] = row.value
# Per-host recent series for the inline sparklines (cpu/mem/load host-level,
# disk from the root mount), so each metric shows its trend beside the value —
# the host-panel at-a-glance, on the fleet widget. 1h window keeps it cheap.
host_names = [h.name for h in hosts.values()]
spark_series: dict[str, dict[str, list]] = {
n: {"cpu": [], "mem": [], "disk": [], "load": []} for n in host_names
}
if host_names:
since = datetime.now(timezone.utc) - timedelta(hours=1)
sp_rows = (await session.execute(
select(PluginMetric).where(
PluginMetric.source_module == SOURCE_MODULE,
PluginMetric.recorded_at >= since,
or_(
and_(PluginMetric.resource_name.in_(host_names),
PluginMetric.metric_name.in_(("cpu_pct", "mem_used_pct", "load_1m"))),
and_(PluginMetric.resource_name.in_([n + ":/" for n in host_names]),
PluginMetric.metric_name == "disk_used_pct"),
),
).order_by(PluginMetric.recorded_at)
)).scalars().all()
for r in sp_rows:
if r.resource_name.endswith(":/"):
name = r.resource_name[:-2]
if name in spark_series:
spark_series[name]["disk"].append(r.value)
else:
key = {"cpu_pct": "cpu", "mem_used_pct": "mem", "load_1m": "load"}.get(r.metric_name)
if key and r.resource_name in spark_series:
spark_series[r.resource_name][key].append(r.value)
from steward.core.status import sparkline_svg
stale_after = _stale_after_seconds()
now = datetime.now(timezone.utc)
rows = []
@@ -359,6 +392,11 @@ async def _fleet_rows(session) -> list[dict]:
m = latest.get(host.name, {})
ls = reg.last_seen_at
stale = ls is None or (now - ls).total_seconds() > stale_after
ser = spark_series.get(host.name, {})
sparks = {
k: (sparkline_svg(v[-60:], width=72, height=16) if len(v) >= 2 else "")
for k, v in ser.items()
}
rows.append({
"host": host,
"reg": reg,
@@ -371,6 +409,7 @@ async def _fleet_rows(session) -> list[dict]:
"temp_max": m.get("temp_c_max"),
"net_rx_bps": m.get("net_rx_bps"),
"net_tx_bps": m.get("net_tx_bps"),
"sparks": sparks,
})
rows.sort(key=lambda r: r["host"].name.lower())
return rows