Three 404s on a quiet router read 7.5% and were raised as an anomaly; most of
the 4xx anomalies on review stood for 3 to 7 errors in 15 minutes. A new
Settings → Thresholds minimum, traefik.threshold.min_4xx (default 20, more
than one a minute held for the window), applies to both the router's grade
on its page and to anomaly detection, as min_5xx does for 5xx.
learn_normal takes `fewest`: for a share gated on a per-second rate, the
fewest events (rate × window × share) it must stand for to be judged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
One 502 among about 70 requests read steward@swarm critical at 1.4%:
past the 30-request minimum and far past the 0.1% cutoff, yet one
error. A new Settings → Thresholds field, "Fewest server errors to
fail" (traefik.threshold.min_5xx, default 3), must also be reached in
the 15-minute window; the 0.1% cutoff still catches a busy router.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Two 500s of about sixteen requests graded roundtable@swarm critical at
9.1%. The Settings → Thresholds minimum of requests in 15 minutes (30 by
default) now guards both shares; the 5xx share still keeps its fixed
cutoff whatever the router's history.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Host CPU, memory and load per core, container CPU and memory, and every
monitor's response time learn a normal per hour (milestone 513, System 182).
A reading past its fixed cutoff but within its normal reads healthy and says
what is usual at this hour; one well out of it keeps its grade and is a
metric anomaly. Disk, a monitor down and an agent offline keep their cutoffs.
- Each normal's floor, the smallest departure that counts, is a setting:
Settings -> Thresholds -> Normals (norms.floor.*). Defaults: host CPU 10
points, host memory 5, load per core 10, container CPU 10 points of one
core, container memory 5, monitor response 100 ms; Traefik's 2-point 4xx
and 50 ms p95 floors move there from constants. learn_normal takes a
floor function (floor_setting) read at each pass.
- learn_normal(parts=False): a host learns, its cores don't.
- Ingest records load_pct (load_1m / cores x 100), the figure pages grade.
- Shared grading: anomalies.against / graded, worst_reading(normals=),
hosts' host_grades, docker usage(thresholds=, norms=), monitors'
latency_grade. Applied to host vitals, the Hosts widget, MCP overview,
Docker usage, and the monitor list, widget and page.
- The monitor and container pages list their anomalies.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5741 "Traefik routers are graded against their own normal", milestone 513.
- A router's 4xx share and p95 learn their normal (learn_normal in the
plugin's setup). Once a router has 5 days of history, a reading over the
fixed cutoff but inside its normal is healthy, and the healthy table says
it is usual for that router. One out of its normal is a warning that says
what is usual, and a metric anomaly.
- The 5xx share is a failure: fixed cutoff always, whatever the history or
traffic (System 182's ruling).
- A 4xx share over fewer requests than Settings -> Thresholds' new "Fewest
requests to judge 4xx" (30 per 15 minutes) isn't graded on the page, and
isn't judged for anomalies.
- Core: learn_normal(gate=(metric, least)) judges a key only while its
resource's gate metric averages at least `least`; `least` may be a
function, so it follows a setting. learn_normal(since=...) learns only
from history after a moment: Traefik latency learns from #5755's
per-scrape figures only.
- The Traefik page lists the routers' metric anomalies of the last 30 days
(_anomaly_card.html). The dashboard widget keeps its why to problems.
- Thresholds and docs say when the cutoffs give way to a normal.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5752 "Container CPU and memory history lives in the one metric store;
docker_metrics and docker_metrics_hourly are retired". Container cpu_pct and
mem_pct already went to plugin_metrics for alerting; now mem_usage_bytes does
too, and every snapshot in an ingest (backlog included) is recorded at its
own time. The container history chart, Busiest Containers, a service's
history and the MCP container tools read plugin_metrics under resource
"<host>/<container>", rolled up and pruned by the shared metric retention.
docker_012_drop_metrics drops both tables. Their history is not copied
(lesson #5221 "A data migration's cost is set by the live table"); the
store already holds 7 days of container CPU and memory %. Docker retention
now prunes only lifecycle events, and Settings -> Thresholds has one
"Metric retention" card for every source.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5740 "A metric that departs from its normal is recorded as an anomaly to
review, and is findable after it ends", milestone 513.
- Every minute each learned key's last 15 minutes is judged against its
hour's normal: above the usual hold x1.25 and by the metric's floor opens
an occurrence, which ends when the level comes back.
- metric_anomalies holds one row per item to review; a repeat while it
still waits joins it as an occurrence (metric_anomaly_spans).
- An ended row nobody reviews within anomalies.review_days (14, Settings ->
Thresholds) lapses; an ongoing one never does.
- Learning skips every hour an occurrence covers unless its anomaly was
marked normal: no normal is learned without review.
- Alerts -> Anomalies lists those to review first, then the reviewed and
lapsed under day dividers; marking is operator/admin and audited.
- The Overview widget and get_overview warn of open ones and count those
to review; the host page shows its own; the weekly report lists the
week's and the lapsed; MCP list_anomalies and review_anomaly.
- Reviewed and lapsed rows are pruned at data.retention_days, and kept
in backups like alert history.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The last page of #5214's walk. Settings still described Steward as it was
before monitors were unified and SNMP was rebuilt, drew its tabs below the
breadcrumb, and showed the report schedule in UTC.
- Tabs (#5377): settings/_tabs.html is a settings_tabs() macro called in
each page's section_tabs block, above the trail, as Ansible's are. The
breadcrumb guard also fails any page that {% include %}s a tab strip.
- Report time (#5378): reports.schedule_tz keeps the zone of whoever saved
the schedule; check_and_send compares in it (report_due), and the form shows
the day and hour on the reader's clock (schedule_in) under their zone's
name. "Last sent" uses |when.
- Wording (#5379): the report contents list what the report holds (monitors,
not ping-enabled hosts); "Ping latency" is "Response time" and says what it
grades; plugin detail says enabling loads now and disabling needs a
restart; the monitor interval says it applies on the next check (the
scheduler reads it every pass); the dead list-config branch is gone.
- Traefik (#5380): the 5xx and 4xx cutoffs are settings
(traefik.threshold.error_5xx_pct 0.1, error_4xx_pct 1.0) in a "Traefik
routers" group on Thresholds while the plugin runs, read by router_state
through error_cutoffs(). The Traefik page has a Thresholds button for
admins, to /settings/thresholds/#traefik.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Agent 1.12.0 stamps each post with sent_at; ingest compares it with the
moment the post arrived: clock_offset_s / clock_checked_at on the
registration (host_agent_003), and a clock_off_secs metric (either way)
that alert rules can watch. Replaces the old skew log, which measured the
newest sample and was off by up to an interval.
- Settings -> Thresholds: Clock drift, 30 s warn / 5 min critical.
- Host page: 'clock 0.3s ahead' in the vitals line, toned past the warning.
- Hosts list: names hosts whose clock is off; when every reporting host is
off by about the same amount, says Steward's own clock is the likely one.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Traefik: routers are graded from their latest scrape (router_state: crit on
any real 5xx share, warn on p95 past the latency critical cutoff or a high
4xx share). Failing and slow routers lead as cards that say why; healthy
ones fold into one table that stays open or shut across polls. Proxy
counts are stat tiles; certificates are graded by Settings -> Thresholds
(tls_days).
UniFi: WAN health is a stat strip (latency graded by the thresholds),
devices carry status dots, event categories and firewall actions are
badges, the alarm count reuses .count-badge (renamed from .nav-count).
SNMP: each device wears a status badge and its readings are stat tiles.
Agent fleet: the table comes first, with a Status column (Reporting /
Silent / Never reported, by the host page's stale_after_seconds), silent
agents first, hosts linked, last seen as |ago.
Audit: entries name their entity (one query per type, steward/audit/
service.py) and link it; something deleted since keeps the name its entry
recorded and says "(deleted)".
Tokens: charts read colours through chart-colors.js (stewardColor resolves
a token to the rgb the browser computed; Chart.js defaults take the page's
text and border colours), so host, history, Traefik and SNMP charts drop
their hex. Template hex colours reach 0 (HEX_CEILING = 0), the last rgba
pills go, and heavy weights (600/700/bold, 60 of them) become 500, the
house pair being 400/500; a guard keeps templates there.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5095 "Share view on the shared stylesheet; breadcrumbs, focus-visible and
wide tables".
- dashboard/share_view.html loads theme.css and steward.css; its own copy of
the old purple palette is gone. Only the read-only frame and the x/y/w/h
grid stay inline, written in tokens. HEX_CEILING 75 -> 58.
- Breadcrumbs on settings/general, settings/thresholds, the dashboard share
page and share-link-created. tests/test_breadcrumbs.py fails on any page
extending base.html without a crumbs() trail, unless it is on the named
list of section top pages, logged-out and error pages (rule 134).
- :focus-visible draws --fs-focus-ring on links, buttons, summaries, inputs
and anything with tabindex.
- Any card holding a .table scrolls sideways (card:has(table.table)), so a
wide table no longer widens the page at phone width; .card-flush scrolls
too, main gets min-width:0, and .table-wrap covers a table outside a card.
- tests/test_share_view.py renders the share view: shared sheets, no hex
colour, escaped name, widgets loading through the share token.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5031 "Forged-request protection: SameSite cookie, per-session CSRF token
on every non-GET, HTMX header". A page on another site could make a
logged-in browser POST to Steward with the session cookie attached.
- steward/auth/csrf.py: a per-session token, checked in one before_request
for every POST/PUT/PATCH/DELETE (header X-CSRF-Token or form field
csrf_token, constant-time compare). The bearer-authenticated host-agent
ingest is exempt; /mcp never reaches Quart.
- Every POST form renders {{ csrf_field() }}; base.html carries the token
in a meta tag and adds it to every HTMX request; the dashboard's fetch()
calls send it.
- The session cookie is SameSite=Lax; login issues a fresh token.
- PUBLIC_BASE_URL is now loaded from Settings at startup (it was only set
when Settings → General was saved).
- tests/test_csrf.py: refusal, acceptance, rotation, the ingest exemption,
and a guard that every POST form in the templates carries the field.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
MCP-5 of milestone 461 "MCP server: Steward as an LLM-inspectable,
LLM-operable hub". Contributed through each plugin's register_mcp_tools
hook, so they exist only while the plugin is enabled.
docker (viewer, read-only):
- list_containers: filter by host, status, compose project, or problem
(restarting, crash_looping, unhealthy, oom_killed, exited_nonzero);
problems sort first and each row names what is wrong.
- get_container: state, CPU/memory history across the raw/hourly seam,
lifecycle events. A unique part of a swarm task name is enough.
- list_swarm_services: replicas per node (agent-less placements as ghosts)
and the swarm's nodes.
- search_container_logs: text (% and _ literal), stream, range, context
lines, paging by a full-precision `before` cursor; over-long lines keep
both ends and say how much of the middle was cut.
- get_docker_disk_usage: bytes per host, reclaimable first.
- mcp_overview: failing containers and under-replicated swarm services
join get_overview.
host_agent: list_agents (reporting / stale / pending, version, OS), and an
overview finding for agents registered over 30 minutes ago that never
checked in.
Settings > Thresholds gains the TLS-expiry warn/crit days (were 14/3 in
code) and the crash-loop deaths-per-hour count (3), which get_overview
and the docker tools now read.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The user-facing half (rule 27). Per-container log viewer + admin controls, so
container logs are something the operator can actually touch.
Viewer (plugins/docker):
- /container/<host>/<name>/logs full page + /logs/lines HTMX fragment polled
every 5s (near-live follow); stream filter (all/stdout/stderr) + case-insensitive
text search; newest-first so the latest lines survive the poll's scroll reset;
empty/loading states; a "View logs" link from the container detail page.
Jinja auto-escapes log content (no markup injection).
Settings (Thresholds & Retention tab):
- global on/off toggle, comma-separated exclude list, retention days + max MB
per container — DB-backed, no restart. The toggle + exclude are enforced
authoritatively at ingest (_persist_logs drops disabled/excluded lines), since
the push model has no channel to tell an agent to stop.
Tests: route-defined smoke; template-parse covers the new templates; integration
test for the ingest-time toggle + exclude enforcement.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CAGR73DUowdVFVvYzLXC5C
plugin_metrics grows by (sources × resources × ~30s cadence); keeping 90d of raw
is a large table. Add a raw→hourly rollup (mirroring the Docker plugin) so only a
short raw window is kept at full resolution, with hourly averages archived longer.
- PluginMetricHourly model + core migration 0024 (plugin_metrics_hourly: avg/max/
count per source/resource/metric/hour, unique bucket constraint + lookup index).
- steward/core/metrics_retention.rollup_plugin_metrics: date_trunc('hour') agg of
raw older than the hour-aligned raw window, idempotent pg upsert into hourly,
delete the rolled raw, prune hourly beyond the rollup window.
- cleanup.py: plugin_metrics is no longer blanket-deleted at data.retention_days;
_run_metrics_retention drives the rollup with windows read live from settings.
- Settings: metrics.retention.raw_days (7) + rollup_days (90), tunable on the
Thresholds & Retention page (new "Host metrics retention" card).
- Chart read: _history_for_host merges the hourly rollup (older part of the range)
with raw date_bin (recent part, capped ≤1h), so 30d charts keep working —
recent at full resolution, older at hourly. Route passes raw_days from settings.
- Tests: unit (cutoff helpers) + integration (rollup aggregates/prunes; history
merges hourly + raw) against Postgres.
Speed was already handled by the indexes + SQL aggregation; this is the storage
lever (raw window ~10x smaller).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC
Bounds Docker time-series growth (the main scaling concern). New
docker_metrics_hourly table + docker_006 migration; a plugin retention module
(docker.run_retention capability) rolls raw docker_metrics older than the raw
window into hourly averages (idempotent upsert), deletes the rolled raw rows,
then prunes stale rollups + lifecycle events. Core cleanup.py drives it each
hourly run via the capability (no plugin-model import), reading the three
retention windows fresh from settings so changes apply without restart (rule 25).
Settings → "Thresholds & Retention" gains a Docker retention card (raw /
rolled-up / events windows, working defaults 7/90/30 days). Unit tests cover the
hour-aligned cutoff/bucketing helpers; integration test exercises the real
rollup-average + prune across both windows.
Milestone 77 task #941.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC
Move the hardcoded warn/crit cutoffs into Settings -> Thresholds (DB-backed,
live, no restart). New thresholds.* keys + to_thresholds_cfg() + a
threshold_style(value, kind) jinja global that reads them; latency reuses the
existing ping good/warn keys, uptime is direction-aware (floors).
Replace the _macros metric_style/uptime_style macros (now removed) with the
global across Hosts-Overview, host_agent fleet + panel, Uptime/SLA widget, and
the ping page uptime column — all now honor the configured cutoffs. Uptime keeps
its green 'good' look when not degraded.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
