Nine services that lost tasks in one node's bad three minutes stayed on the
overview as warnings for a day, 17 hours after they recovered. Against the
UI ruling (only what is current on the dashboard), a service running all its
replicas is now history: its failed tasks stay on the Docker page, not on
the overview or the dashboard's failed count. Services still short of
replicas that started failing within five minutes of each other read as one
finding naming them, dated from the first.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
VDNT-Docker03's silent agent read as three overview findings (agent offline,
agent stale, the disk rule's no-data) and every drop emailed the disk rule's
silence and return on top of agent offline.
- The host agent registers its heartbeat (seconds_since_seen) and the
sources it carries (host_agent, docker). While a host's agent is past its
offline window, its other rules' no_data is recorded "while its agent is
offline" and not emailed, nor is their quiet return; a return past a
threshold still fires.
- The overview folds those rules under the host's agent-offline finding
("also silent with it: ..."), and a stale agent whose agent-offline alert
is active is that alert's finding, not a second one.
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>
#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>
#5468 "Log search by service name: searching "steward_app" refuses because every
deploy makes a new container name".
search_container_logs(container="steward_app") was refused, because no container
carries the service's own name and five task containers partly matched. Now a
service's name reaches every task container Steward holds anything for, live and
removed (docker_svc.service_containers, as the service page's logs already do),
so one search spans deploys. Each match carries the slot and task_id that wrote
it, and the reply names the service and how many tasks were searched. A container
named exactly what was asked still takes precedence. A miss that partly matches
task containers now names their services.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5449 "The hourly Docker log prune ranks every log line in the table to find
the few over their cap": the size ring put a running-total window over every
row of docker_logs each hour, sorting the whole table on NFS-backed storage
to find the handful of containers past docker.logs.max_bytes_per_container.
- prune_docker_logs (capability docker.prune_logs) does the age cutoff, then
a per-container byte total picks the containers at or over the cap, and
only their rows are ranked and trimmed. Same keep rule as before: the
newest line always survives.
- run_docker_retention keeps the metrics rollup and the event prune.
- Cleanup runs the log ring in its own transaction after them.
Arose from #5431 "Ingest dropped for ~3 minutes after each deploy while
startup retention runs".
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Confirm dialog (#5314):
- Every data-confirm form now asks in one native <dialog> (base.html + static/confirm.js)
instead of window.confirm.
- The question becomes the heading and what follows it the detail; the act's
button is named after the form's own button; Cancel has the focus and Esc
backs out. Optional data-confirm-title/-detail/-action/-dismiss attributes.
- All 27 existing uses work as they are. Cancelling a run says 'Keep running'.
Alerts list (#5315):
- A verdict line (what is firing, how many rules, where notifications go,
or 'notifies nobody yet' with a link) replaces the empty Active card.
- Each rule's name opens it; rows show 'last fired' and fold fire/clear into
the condition; History and Disable stay; Delete leaves the list.
- .docker-verdict becomes the shared .page-verdict.
Rule page (#5317):
- Named for the rule (title and crumb), with when it last fired and how many
changes of state it holds.
- Three sections: what it watches, when it fires (checks side by side,
seconds echoed as a duration), what happens (where notifications go, no-data,
the playbook, whose fields stay folded until ticked).
- Delete at the foot, behind the dialog, saying its history goes with it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Service page:
- The header carries the verdict, mode, stack, image and when it was reported.
'In trouble since' shows only while the service is in trouble; once recovered,
'Last failed X ago · its logs'.
- The failure card shows only while the service is down, degraded or crash-looping.
- The tile strip and the kv card are gone.
- One Tasks table: running tasks first with CPU, memory and since when, then
Swarm's record. The node is named as its Steward host throughout, and
'N failed in 24h' sits in the caption.
Container page:
- Same shape: header verdict, a 'Last exit' card only while it is stopped
after a failing exit, set-up and I/O in one kv card, and lifecycle as a table.
- A swarm task's breadcrumb goes through its service.
Both pages:
- Resource history is the one fragment (_usage_history.html): now and peak in
cores and bytes, drawn with a floor of 1 core / 1 GiB (sparkline_svg floor=),
so idle noise draws flat.
- The container page's old % sparklines, its private _sparkline and the unused
bucket()/container_history() helpers are removed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- #5304 "Docker disk totals recorded over time per host, with a growth chart and alertable metrics"
- #5305 "Disk page: images grouped by repository, versions folded under it"
- #5306 "Docker overview: one disk line per host linking to the Disk page"
Each `docker system df` report is now also recorded as metrics, with the host
as the resource: disk_total_bytes, its parts, and disk_reclaimable_bytes. An
alert rule can watch them. The Disk page starts with the hosts side by side,
showing how much each grew over the selected range with a line of it, then
gives each host its own section. An image's versions fold under its
repository, which is charged only for the layers its versions don't share.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5303 "Swarm page laid out by node: role, resources and the tasks placed on each"
A verdict line first, then the nodes that are down or drained and the
services short of replicas, then one card per node. A node with a Steward
agent shows what its swarm tasks use against the machine's cores and memory,
service by service. A node without one lists Swarm's placement, by count.
The services table moved to the Docker overview, where it can be filtered.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- #5298 "Docker overview: verdict line; only the problem tiles that aren't zero"
- #5299 "Docker overview: swarm services as a filterable, sortable table with a Stack column"
- #5300 "Docker overview: failures listed once, each with a link to the failed task's logs"
- #5301 "Docker overview: standalone containers in one table with a Host column"
- #5302 "Service page: replicas, image, task history, resource history and logs across all its tasks"
overview.py holds the rows, filters and sort as plain functions. The filter
bar sits outside the polled fragment and the URL keeps its state. A service's
page reads every task container Steward still holds anything for, the live
ones and removed ones known by their events, so a failed task's logs are
reachable after Docker has removed it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Docker panel (#5276): problems first, then the 8 busiest running
containers by CPU; the other running and the stopped fold behind one
"+N more running · M stopped" line. Containers that are not swarm tasks
read the way the Docker page reads them: CI jobs as "CI: workflow · job",
buildx builders as "BuildKit builder · name".
Monitors card (#5277): the edit/delete icon buttons and the inline add
form are gone. Each row links to its monitor's page; "Add monitor" opens
the add form filed under this host, starting from its address, with the
host in its breadcrumbs and Cancel back to it.
Agent card (#5278): Full metrics stays in view; Update, Rotate token,
Unregister agent (was "Remove") and Re-provision sit under "Manage agent".
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Each figure after its bars takes a fixed width, so memory figures of
different lengths no longer shift their bars. CPU and Memory headers say
"last 6h" and the note gives the slice length.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The fill bars were a share of the host's capacity, a sliver on hosts with
80 cores. Each figure now sits beside 24 bars of the last 6 hours (1h/24h
per widget), each row scaled to its own peak but never below 1 core or
1 GiB (configurable per widget), so a 0.05-core container reads flat and
a bar's height means real load. One grouped date_bin query feeds them.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
View mode splits each saved row into seven, so a widget no longer carries up
to 69px of empty space below its content (Overview, SNMP). Edit mode and the
saved layout keep 70px rows. The Busiest Containers host column aligns left
like its header, its note fits one line, and the Hosts column reads Disk.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The widget names the busiest container in a verdict ("Busiest: tdarr on
VDNT-Docker03, 1.6 of 8 cores."), then a fixed-layout table (container,
host, CPU in cores, memory in bytes) with each figure's share drawn as
shading behind it: of the host's cores, and of the memory limit or the
host's memory.
The numbers were already rendered, but each row grew as wide as its longest
name. A grid item won't shrink below its content, so Gitea CI container
names pushed the figures out of view and the cell scrolled sideways. Names
now come from display_name(): a swarm task by its service (no slot number),
an Actions job container as "CI: <workflow> · <job>". They cut off with …
and show whole on hover.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A verdict line (services healthy · other containers running · failed in
24h), then "Needs a look": troubled services with why (crash-looping, N of
M running), since when and the last error, and troubled containers
(unhealthy, restarting, exited with an error, out of memory). Then "Failed
in the last 24h" by name, with how often, when and where. Then the healthy
rest one line each (name · hosts · replicas) in as many columns as the
widget is wide, collapsing after 40 to "+N more healthy".
Per-container CPU is gone from this widget: Busiest Containers owns it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5212 "Plain labels: no raw ids, scopes, metric keys or unexplained jargon",
batch 3:
- Disk page: each tile says what it holds. "Writable layers" becomes
"Container files" and Build cache says it is safe to clear. Shared is
"Shared with other images" with a tooltip, and "dangling" is gone.
Untagged images show by id instead of "<none>", a capped list says
"The 50 largest of N images", and each host shows "as of" its report.
- Traefik: an expired certificate reads "expired 12 days ago", not "-12
days" (cert_left). The Traffic empty state knows whether the access log
is on: off points to Settings → Plugins → Traefik; on says to check the
mounted log path.
- Moments use |ago with an exact UTC tooltip: UniFi last seen, alarms,
events, speed test and DPI; AI token last used.
- Settings: the Auth tab is "Authentication" like its breadcrumb, plugin
fields read "Metrics URL" not metrics_url (field_label), a source's pull
interval is "every 1h", and the last report time drops the ISO "T".
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5212 "Plain labels: no raw ids, scopes, metric keys or unexplained jargon",
batch 2:
- A run records where it came from: AnsibleRun.origin ("schedule", "alert"
or "mcp") and origin_id, set by the scheduler, Run now, alert actions and
the MCP tools. Migration 0031 backfills past runs from alert events,
schedules' last runs, and MCP audit entries.
- The runs list has a Started by column: "schedule nightly", "alert rule
nas down", "alice via MCP" or "alice", linked where there is a page.
The raw scope tooltip is gone.
- Run detail: no run id in the subtitle or body, Started by in words, and
the host summary lists failed and unreachable hosts first. The live
output only follows new lines while the reader is at the bottom, and
when the run finishes the page reloads once so the badge and summary
are final.
- list_ansible_runs and get_ansible_run report origin and origin_id.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
failing_services filtered on `(service_trouble(s) or 9) <= 2`; a crash loop
ranks 0, which is falsy, so the worst kind of trouble was left out. Test the
rank against None. The page test's task error also gains the age_s the
_swarm_failures include reads.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5209 "Docker views: readable names, CPU/memory in cores and bytes, failures
listed with since-time" (milestone 477 "UI problems first").
- A swarm task reads as its service and slot ("codex_db · 1", or the node
for a global service), with the raw name in the title: Docker page, host
panel, busiest widget, container detail.
- CPU reads in cores (Docker's 100% = one core), toned by its share of the
host's CPUs from the agent's per-core series. Memory reads in bytes, against
a limit only when one is really set (Docker reports host RAM otherwise).
- Exit 137/143 without an OOM kill is a stop, not a failure: no
exited_nonzero, no crash-loop death, no "failed (24h)", dim exit code.
- "Failed (24h)" is a list (Docker page #failures, linked from the widget
and the tile): plain containers by their deaths, swarm services by Swarm's
failed tasks; a second manager's copy of a task death counts once.
- Failing swarm services lead Needs attention; services sort problems first
with since-times on the page, the widget and the Swarm page; nodes that
are down or drained sort first.
- Busiest widget orders by a 5-minute CPU average so the list holds still.
- Host panel: a table, problems first with what is wrong and since when;
replaced swarm tasks hidden and counted.
- Container detail: node hostname, I/O labelled "since start", Restarts
hidden for swarm tasks with a note on where task failures are.
- Widget names: "Docker — Services & failures", "Docker — Busiest containers".
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Agent 1.11.0 sends finished_at (State.FinishedAt) and reads a stopped
container's logs once per stop: since its cursor if it was being read, or
its last 200 lines when first seen already crashed (non-zero exit, stopped
in the last 10 min). Cursors and finals are forgotten once a container is no
longer listed.
Ingest gives a container first seen already exited non-zero a die event, and
dates die/stop events by finished_at when the agent sends it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
persist_host_docker marks docker.status, .swarm, .disk, .logs, .metrics,
.events, .containers, .alerts and .reap on the ingest's phase timer, each
flushed before its mark so a part's writes are charged to it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Measured on dev (385a579), agent posts cut to five samples still dropped:
flush=8.8s for 990 rows, 6.6s for 680, 8.4s for 1000 — about 110 rows/s,
~9 ms a row, with rules and evaluation near zero. One random read per
row fits a random uuid4 primary key whose index no longer fits in memory:
every new row lands on a random page of it. (Inferred from the rate; the
table and index sizes have not been read.)
- steward.core.ids.time_ordered_id: a UUIDv7-layout string (48-bit ms
timestamp, then random), same 36-char shape as the stored uuid4 ids, so
no migration. Python's uuid7() arrives in 3.14; the image runs 3.13.
- Used for plugin_metrics and its hourly rollup, monitor_results,
audit_events, the docker metrics/hourly/events/logs tables, and the
traefik and unifi time series.
- MAX_SAMPLES_PER_INGEST 5 → 2 (~400 rows) for now: it fits the agent's
10 s even at the old rate, so hosts land data whichever way the new rate
comes out. Raise it once the new rate is measured.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Agent: started_at comes from inspect's State.StartedAt (nanoseconds
trimmed; Docker's zero time is "never started"), not the list's Created,
so a restarted container's uptime counts from its restart.
- A rising restart_count between two snapshots emits a "restart" event
("restart #3", or "3 restarts, now #5"), said instead of a start; a
restart then a death gives both. A count that falls (recreated under the
same name) says nothing. The timeline draws it as ↻.
- Network and disk I/O count what the kernel calls physical: interfaces
not under /sys/devices/virtual/net (bridges, docker0, veth, vxlan, tun),
block devices with a device behind them (not dm, md, zram), so a byte is
not counted twice. Where /sys calls everything virtual (an agent in LXC)
the name-filtered set is kept rather than reporting no traffic.
- Agent tests updated for #5152's (status, detail, containers) and
(is_manager, topology) shapes; the no-Docker disk test now states the
case it was passing for by accident.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The agent (1.9.0) said nothing when Docker could not be asked and nothing
when Docker had no containers, so Steward could tell neither apart from the
other, and kept a host's last containers on screen as running either way.
- Agent: collect_docker returns (status, detail, containers). "ok" sends
docker: [] when there are none; "unavailable" (no socket, nothing
listening) and "error" (permission, timeout, bad answer) say why.
collect_swarm says whether /info answered, and a node that is not a
manager sends swarm_manager: false. A failed /info says nothing.
- Ingest: an empty docker list is a snapshot; the newest sample's status
goes to the Docker plugin with the rest.
- docker_host_status (docker_011): per host, status, detail, since and
last report. A host that never had Docker gets no row.
- An "ok" empty snapshot reaps the host's containers; an error keeps them.
swarm_manager: false drops the host's swarm services and nodes.
- Docker page: a "Docker not answering" card names each such host, why,
and since when; its containers read as stale. Duplicate header comment
in rows.html removed.
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>
The Docker page opens with a "Needs attention" card: every container
with something wrong, worst problem first — restarting, crash-looping,
unhealthy, OOM-killed, exited non-zero — each with what it does ("is
restarting and keeps dying · exit 1 · since 8m ago"), linked to its page.
It is ranked by the same code as the MCP tools: container_problems,
problem_since and the per-container ranking move from mcp_tools into
service.py (ranked_containers, needs_attention), and mcp_tools' _rows and
findings now read them from there.
Counts follow as stat tiles: needs attention, unhealthy, restarting (or
crash-looping), running, stopped, stale, services, hosts.
A container whose last report is older than the host agent's staleness
window (stale_after_seconds, 180 s by default) carries a Stale badge with
when it last reported, instead of a live-looking green dot.
The per-container CPU and memory history columns are gone from the
overview, and with them one history query per container per refresh; the
container's own page keeps its charts. With nothing left to range over,
the time-range control goes too, and the page now refreshes on the poll
interval. CPU and memory are graded by Settings → Thresholds rather than
50/80 and 70/90 written into the template.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5099 "A name containing a quote runs script when an admin clicks Delete".
Inline handlers like onclick="return confirm('Delete {{ rule.name }}?')" are
HTML-decoded before the JavaScript is parsed, so autoescaping did not stop a
name such as x');fetch(…);// from running.
- Every prompt (22 handlers, plus the docker prune and token revoke ones)
is now data-confirm="…", read through dataset by one submit listener in
base.html; nothing rendered is parsed as JavaScript.
- The remaining inline handlers take integers only (|int), and the time
range buttons pass this.dataset.range.
- tests/test_inline_handlers.py fails on any {{ }} in an inline handler that
is not |int, with a self-test; tests/_templates.py is the one template
walker the template guards now share.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5093 "Shared format filters and status_class driven by Settings → Thresholds".
- steward/core/formatting.py: human_bytes (binary units), human_rate,
human_duration (two largest units), ago (a <time> titled with the absolute
UTC time) and status_class(value, kind, thresholds) -> ok/warn/crit/unknown,
which raises on a kind the thresholds don't name.
- app.py registers them as the bytes, rate, duration and ago filters and the
status_class global; threshold_style is now graded by status_class (and caps
its weight at 500).
- Replaces the copies: docker's _human_bytes/_human_uptime, core alerts'
_duration, the host-agent fmt_bps/fmt_bytes macros, the Traefik and UniFi
byte-rate arithmetic, the swarm-failures ago macro and two uptime spans.
- Tests: a table per helper, and every template compiled against the app's
filters, so an unknown filter fails CI rather than a page.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5061 "Metric catalog belongs to what emits: plugins register their
metrics; names match what is recorded".
- The static METRIC_CATALOG is gone. metric_catalog() is core's own sources
(the four monitor types and "steward", the app's open_fds/open_fds_pct,
which were recorded but never offered) plus each loaded plugin's
get_alert_metrics() export. A disabled plugin's sources leave the picker.
- traefik offers what it records: error_rate_4xx_pct and error_rate_5xx_pct.
The old "error_rate" was never recorded, so a rule on it could not fire.
- tests/core/test_metric_catalog.py walks every MetricBatch.add /
record_metric call (traefik's tuple loop included) and fails on a recorded
metric the catalog lacks; it shows on a synthetic writer that it can fail.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5060 "One evaluation path: record_metrics batch API; host-agent ingest
evaluates rules".
- record_metrics(session, rows) writes metric rows and evaluates the rules
that watch them, reading rules, states and maintenance windows once per
batch. Each rule sees its key's samples oldest first, so an agent catching
up after an outage counts consecutive checks as they happened.
- Host-agent ingest now goes through it. Before, it wrote plugin_metrics
directly and no host_agent rule could ever fire. The catalog gains the
per-mount disk_used_pct ("<host>:<mount>").
- MetricBatch gathers a transaction's samples; the docker ingest, traefik,
unifi, snmp, monitor and self-monitor writers record one batch each instead
of a rule query per sample.
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>
#5010 "Logs of a removed container cannot be reached by name"
- get_container and search_container_logs fall back to the names Steward
still holds logs or events for when no live container matches: the exact
name from logs and events, or a part of it that matches one removed
container's events.
- get_container answers with removed: true, last_seen_at, its history and
events; search_container_logs marks container_removed and notes when it
was last seen, and says when that is older than the range.
- A miss now says the container may have been removed, that Swarm replaces a
dead task under a new name, and where to look instead.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
#5013 "Overview findings for containers and Swarm services carry since=null"
- Container problems are dated from their own evidence: a crash or restart
loop from its first death in the hour, an OOM kill or error exit from the
death that ended it, a failing healthcheck from its last turn to unhealthy.
list_containers and get_container carry problem_since / problem_age_s.
- Swarm services carry trouble_since: a crash loop's first failure in the
hour, a short service's oldest troubled task in each failing slot, a
recovered service's first failure in 24h. list_swarm_services shows it.
- get_overview findings from the docker plugin now have since and age_s,
so they rank by how long they have stood.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Milestone 465 "Swarm task awareness". Fixes three defects from the live MCP
review #4973:
- #5008 "Swarm task history read as live state"
- #5009 "Crash-loop detection is blind to Swarm services"
- #5011 "Swarm task errors are not captured"
Agent and ingest (agent 1.8.0):
- A manager's agent now reports each service's newest tasks: slot, node,
desired and current state, Swarm's Err/Message, exit code and time.
- They are stored in docker_swarm_services.tasks_json (migration
docker_010_swarm_tasks).
Swarm view (pure):
- A task container is "replaced" when Swarm set its task's desired state off
`running`. Without the task on record, a newer or running task in the same
slot decides.
- Replaced containers are no longer replicas or crashes.
- Each service gets, from the task records:
- failures_1h and failures_24h (failed or rejected tasks);
- crash_looping (Settings → Thresholds count);
- last_error, in Swarm's own words.
MCP:
- New findings: swarm_service_crash_looping (critical) and swarm_task_failed
(warning, on a healthy service with a failure in 24h).
- Down and degraded findings name the last task error.
- list_swarm_services carries the failure fields.
- Container rows mark swarm_replaced and give a replaced task no problems.
Web:
- The containers page swarm cards and the Swarm topology page show crash
loops, recent failures and the last task error.
- The dashboard widget flags a crash loop, and its "failed (24h)" count no
longer counts a rolling update's killed task.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
MCP-6 #4970 "Automation tools: playbooks, runs, schedules, alert actions".
Viewer: list_playbooks, get_playbook (secret defaults masked), list_inventory
(variable names only), list_ansible_runs, get_ansible_run (log tail + grep),
list_schedules, wait_for_run (polls, 120s cap).
Operator: run_playbook (check mode unless check=false; a steward:confirm
playbook applies only with confirmed=true; secret variables reach the run
but are never stored, audited or returned), cancel_ansible_run, and
create/update/delete_schedule (secrets refused, check by default).
Admin: set_alert_action, and the docker plugin's prune_docker, which
previews from the disk report unless check=false (Ansible's check mode
skips the prune commands, so a check run would show nothing).
Every write is audited as "<user> via MCP" with the token id; the audit
tool's user filter includes those rows. prune_extra_vars moves to the
docker service so the page and the tool share it.
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>
MCP-3 of milestone 461 "MCP server: Steward as an LLM-inspectable,
LLM-operable hub". The /mcp tools run outside Quart's request pipeline, so
the queries they share with the pages now live in plain functions that take
an open session:
- steward/hosts/service.py: uptime, status, monitors, agent metrics,
agent overview (stale threshold passed in, not read from current_app),
CPU sparks.
- plugins/docker/service.py: bucket, container_history, host_map,
list_containers, container_events, query_logs.
- steward/alerts/service.py: active_alerts (also used by core/reports),
list_rules, list_maintenance_windows.
- steward/monitors/service.py: monitor_rows (was routes._rows_data).
Routes are repointed and render the same data. One behavior fix rides
along: the container-log search now escapes % and _ so they match
literally instead of acting as LIKE wildcards.
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
Bound docker_logs growth in the periodic cleanup task (same architecture as the
metrics/events retention). run_docker_retention gains logs_retention_days +
logs_max_bytes_per_container: it prunes lines past the age window, then keeps
only the newest ~cap bytes per (host, container) via a window-function ring
(exclusive-prefix sum, so the newest line always survives even if it alone
exceeds the cap). Containers rotate independently.
- settings DEFAULTS: docker.logs.enabled/exclude/retention_days(3)/
max_bytes_per_container(5MB) — operator preference: ~3 days / ~5 MB
- cleanup.py reads the two windows fresh each run (rule 25, no restart)
- integration rotate test: age prune + per-container byte cap + isolation
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CAGR73DUowdVFVvYzLXC5C
Ingest side of container logs. The host_agent ingest route extracts
sample["docker_logs"] as time-series batches (append every line, each carrying
its own Docker ts with recorded_at as fallback) and hands them to the docker
capability alongside the existing container/swarm/disk data — still under the
begin_nested SAVEPOINT so a logs failure can't sink host metrics.
- ingest.py: _persist_logs + pure _log_rows(batches, host_id) that shapes/filters
records (drops the _steward truncation marker, malformed + lineless records,
normalises unknown stream → stdout); persist_host_docker gains a logs= param
- routes.py: accumulate docker_log_batches, add to the guard + capability call
- unit test for _log_rows shaping/filtering; integration push→store→query test
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CAGR73DUowdVFVvYzLXC5C
Dedicated host-scoped table for pushed container log lines (one row per line),
chained after docker_008. host_id FK CASCADE; twin indexes — (host_id,
container_name, ts) for the viewer, ts alone for the age-cutoff prune — mirroring
docker_events. Integration schema-shape test asserts columns + both indexes.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CAGR73DUowdVFVvYzLXC5C
M78 step 2. The /disk page surfaced reclaimable space + stopped counts but
said "prune deferred to a later milestone" — this wires the cleanup.
Adds POST /plugins/docker/disk/<host_id>/prune (admin only) that resolves
the host's linked AnsibleTarget → steward:target:<id> scope and fires the
bundled maintenance/docker_prune.yml through the ansible.run_playbook
capability (audited runner; the collection agent stays read-only). Mirrors
the host_agent update route. `target` ∈ {containers, images, system}:
- containers → docker container prune -f
- images → docker image prune -af (prune_all_images)
- system → docker system prune -f (conservative)
The disk() route now resolves each host group's linked target + capability
availability; disk.html renders three confirm-gated buttons per host, shown
only to admins when Ansible is available and the host has a linked target,
otherwise a clear "link a target" / "runner unavailable" hint (rules 24/26/27).
Tests: the prune_target→extra_vars mapping (the -af-only-for-images rule) is
extracted to a pure helper and unit-tested; disk_prune added to the module
smoke; disk.html covered by the existing template-parse test.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CAGR73DUowdVFVvYzLXC5C
The container list and dashboard widget listed each swarm task as its own cryptic
row (svc.1.<taskid>) and could only show tasks on nodes that run a Steward agent.
Make both views service-centric and cluster-complete.
- swarm_view.build_swarm_services(): model-free builder that merges real container
rows (collected local-per-node, so host_id = the node a task runs on) with the
managers' placement (DockerSwarmService.placement_json). Where placement counts
exceed the local rows on a node — i.e. nodes with no agent — it synthesizes
"ghost" replicas so the replica list matches the cluster. Non-swarm containers
pass through grouped by host.
- /rows + rows.html: a Swarm-services panel collapses each service to one block,
every replica a host chip (status + cpu); ghosts render dashed "N · no agent".
Non-swarm/compose containers keep the per-host table below. New Services stat.
- /widget + widget.html: same service collapse with host chips; standalone
containers stay grouped by host.
- Unit tests for the builder (real+ghost merge, full coverage, partial agent-less,
synthesized service, down state).
No migration — node_id, placement_json and swarm-node hostnames are already stored.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC
The widget-dedup test imported plugins.docker.routes, which re-imports
plugins.docker.models after the app already loaded the docker plugin —
"Table 'docker_containers' is already defined". Same plugin-loader gotcha the
host_agent query helpers avoid.
Extract dedup_by_container_id into plugins/docker/dedup.py (imports no ORM
models, so it's safe to import in either environment); routes.py imports it
from there and the test targets the model-free module.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC
The running count was inflated: swarm-aware agents on every manager list
cluster-wide tasks, so the same container (identical container_id) was reported
once per manager and the dashboard widgets summed them. The earlier swarm dedup
only covered the Swarm topology page — the widgets read raw rows.
- _dedup_by_container_id(): collapse rows sharing a non-empty container_id
(globally unique, so only true duplicates merge); first occurrence wins under
the running-first ordering. Applied in both the containers and resources
widgets before counting/limiting. Rows without a container_id (older agents)
are kept as-is.
- Replaced the static "stopped" tally — not actionable — with "failed (24h)":
distinct containers with a die/oom event in the last 24h (DISTINCT so the
managers' duplicate events don't double it), rendered red when non-zero.
- widget.html empty-state now keys off total_count.
- Regression test: same container_id from two managers counts once; id-less rows
all kept.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC
The container persister upserted current state keyed (host_id, name) but — unlike
the image/swarm/disk persisters — never deleted containers that vanished from the
host's latest listing. Every removed one-shot container (CI job runners, buildkit
builders, codex jobs) left a permanent "stopped" row, so the dashboard counts
ballooned (e.g. 856 stopped) and read as "not dedup'd". It wasn't dedup — it was
a missing reaper.
- ingest.py: after the upsert loop, delete this host's containers whose name is
notin the newest snapshot (the listing is authoritative — event derivation
already treats absence as removal). Mirrors images/swarm/disk. The existing
`if not snapshots: return` keeps swarm/disk-only samples from touching
containers; an empty container snapshot legitimately means none exist.
- widget_resources.html: this widget shows the busiest running containers by CPU.
Replaced the two unlabeled hairline bars with labeled CPU/MEM bars (label · bar ·
% value, warn/crit colors) so it's self-explanatory.
- Regression test: a container present then absent across snapshots is reaped.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC
