feat(prometheus): the Prometheus dashboard widget, target + template (#5691)
CI / lint (push) Successful in 3s
CI / integration (push) Successful in 1m14s
CI / unit (push) Successful in 1m29s
CI / publish (push) Successful in 1m12s

#5691 "Generic Prometheus dashboard widget: pick a target and a template",
step 3 of milestone 508 "Prometheus widget templates".

- /plugins/prometheus/widget draws one target with one template in widget
  shape: the verdict (led by "isn't answering" when the target is down),
  then each block with problems first, figures beside a share bar or the
  last 24h of bars, and the healthy rest folding to "+N more healthy" past
  six. A missing target or template, or one that no longer checks, says so.
- library.py: bundled templates from plugins/prometheus/widgets/*.yaml,
  the picker's target and template lists, and the default title.
- widgets/fabled_archives.yaml: the first bundled template, against the
  series Archives reports (refined in #5693).
- Dashboard: a widget setting may take `options_from` (read when the picker
  opens) with an `empty` sentence; `default_title`; a Details link may name
  settings ("/plugins/prometheus/targets/{target}"). Prometheus is grouped
  with the integrations.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
2026-10-14 07:05:15 -04:00
co-authored by Claude Opus 5.5
parent 4d95c5d0c2
commit 5ca13f9278
11 changed files with 551 additions and 8 deletions
@@ -0,0 +1,99 @@
"""Integration: the Prometheus dashboard widget (#5691) draws a target's
stored series with the bundled Fabled Archives template, the picker lists the
targets and templates, and a widget whose target is gone says so.
Requires STEWARD_DATABASE_URL."""
from __future__ import annotations
import asyncio
import importlib
import os
import time
import uuid
import pytest
pytestmark = pytest.mark.integration
@pytest.fixture
def app():
if not os.environ.get("STEWARD_DATABASE_URL"):
pytest.skip("needs Postgres")
from steward.app import create_app
return create_app(testing=False)
def _prom(sub):
"""A submodule of the prometheus plugin as the app loaded it (see
test_prometheus_scrape)."""
from steward.core.plugin_manager import get_loaded_plugin_modules
loaded = get_loaded_plugin_modules().get("prometheus")
return importlib.import_module(f"{loaded.__name__ if loaded else 'plugins.prometheus'}.{sub}")
def _series(name: str) -> list[tuple[str, str, float]]:
"""(resource, metric, value) as a scrape of Archives stores them."""
rows = [(name, "fabledarchives_data_disk_free_bytes", 17.8e12),
(name, "fabledarchives_data_disk_size_bytes", 46e12),
(name, "fabledarchives_files", 94116.0), (name, "fabledarchives_files_checked", 93310.0),
(name, "fabledarchives_malware_held_files", 2.0),
(name, "fabledarchives_notification_channels_failing", 0.0),
(f"{name}:version=2026.10.12.1717,channel=dev,commit=3605d0a", "fabledarchives_build_info", 1.0),
(f"{name}:person=Bryan,person_id=9,kind=files", "fabledarchives_person_stored_bytes", 4.7e11)]
for state in ("disabled", "fresh", "late", "missing", "waiting"):
rows.append((f"{name}:target=Steward,target_id=1,state={state}", "fabledarchives_backup_target_state",
float(state == "late")))
for state in ("ok", "failing", "late", "running"):
rows.append((f"{name}:job=Nightly,job_id=5,state={state}", "fabledarchives_replication_job_state",
float(state == "ok")))
return rows
def test_the_widget_draws_a_target_with_a_template_and_the_picker_lists_them(app):
from steward.core.alerts import record_metrics
from steward.core.widgets import WIDGET_REGISTRY
from steward.dashboard.routes import with_options
from steward.models.metrics import PluginMetric
from steward.models.users import User, UserRole
PrometheusTarget = _prom("models").PrometheusTarget
tag = uuid.uuid4().hex[:8]
name = f"arch-{tag}"
async def go():
uid = str(uuid.uuid4())
async with app.db_sessionmaker() as s:
async with s.begin():
s.add(User(id=uid, username=f"pw-{tag}", email=f"pw-{tag}@test", password_hash="x",
role=UserRole.viewer, is_active=True))
target = PrometheusTarget(name=name, url=f"http://{name}/metrics")
s.add(target)
await s.flush()
target_id = target.id
await record_metrics(s, [PluginMetric(source_module="prometheus", resource_name=r,
metric_name=m, value=v) for r, m, v in _series(name)])
async with app.db_sessionmaker() as db:
picker = await with_options(db, WIDGET_REGISTRY["prometheus"])
client = app.test_client()
async with client.session_transaction() as sess:
sess.update(user_id=uid, user_role="viewer", username=f"pw-{tag}", login_at=time.time(), tz="UTC")
ok = await client.get(f"/plugins/prometheus/widget?target={target_id}&template=bundled:fabled_archives")
page = " ".join((await ok.get_data(as_text=True)).split())
gone = await client.get("/plugins/prometheus/widget?target=no-such-target&template=bundled:fabled_archives")
async with app.db_sessionmaker() as s:
async with s.begin():
await s.delete(await s.get(PrometheusTarget, target_id))
return target_id, picker, ok.status_code, page, gone.status_code, await gone.get_data(as_text=True)
target_id, picker, status, page, gone_status, gone = asyncio.run(go())
targets, templates = picker["params"]
assert {"value": target_id, "label": name} in targets["options"]
assert {"value": "bundled:fabled_archives", "label": "Fabled Archives (built in)"} in templates["options"]
assert status == 200
# The late backup target and the held malware are the problems, the late target first.
assert "2 need a look. Backup targets: Steward is late." in page
assert page.index("2 need a look") < page.index("Backup targets</div>") < page.index("Replication</div>")
assert "16.2 TiB of 41.8 TiB" in page and "93,310 of 94,116" in page
assert "version 2026.10.12.1717 · channel dev · commit 3605d0a" in page
assert f'<a href="/plugins/prometheus/targets/{target_id}">{name}</a>' in page
assert gone_status == 200 and "The target this widget showed has been deleted." in gone
+122
View File
@@ -0,0 +1,122 @@
"""The Prometheus dashboard widget (#5691): one target drawn by one template,
in widget shape, and the dashboard pieces it needs (settings read when the
picker opens, a Details link naming the target)."""
import asyncio
from datetime import timedelta
from plugins.prometheus import library
from plugins.prometheus.routes import spark_heights
from plugins.prometheus.template_format import check
from plugins.prometheus.template_view import render
from steward.dashboard.routes import detail_url, with_options
from tests._templates import render_plugin_template
from tests.plugins.prometheus.test_templates import NOW, SERIES, TEMPLATE, T, latest, states
from tests.test_style_tokens import hex_colours
def test_every_bundled_template_is_a_template():
entries = library.bundled()
assert [e.key for e in entries] == ["bundled:fabled_archives"]
for e in entries:
assert e.template is not None, [str(p) for p in e.problems]
assert entries[0].name == "Fabled Archives" and entries[0].template.expects == "fabledarchives_"
def test_settings_read_when_the_picker_opens():
async def targets(db):
return [{"value": "t1", "label": "Archives"}]
async def none(db):
return []
widget = {"key": "w", "params": [
{"key": "target", "type": "select", "default": "", "options_from": targets},
{"key": "template", "type": "select", "default": "", "options_from": none, "empty": "No templates yet."},
{"key": "hours", "type": "select", "default": 6, "options": [{"value": 6, "label": "6h"}]}]}
filled = asyncio.run(with_options(None, widget))
assert filled["params"][0]["options"] == [{"value": "t1", "label": "Archives"}]
assert filled["params"][1]["options"] == []
assert filled["params"][2] is widget["params"][2]
assert "options" not in widget["params"][0] # the registry isn't changed
plain = {"key": "x", "params": []}
assert asyncio.run(with_options(None, plain)) is plain
def test_details_names_the_target_or_falls_back_to_the_section():
defn = {"detail_url": "/plugins/prometheus/targets/{target}"}
assert detail_url(defn, {"target": "a b/c", "template": "x"}) == "/plugins/prometheus/targets/a%20b%2Fc"
assert detail_url(defn, {}) == "/plugins/prometheus/"
assert detail_url({"detail_url": "/monitors/"}, {"limit": 5}) == "/monitors/"
def test_bars_scale_to_their_own_peak():
assert spark_heights([1.0, 2.0, 4.0]) == [25.0, 50.0, 100.0]
assert spark_heights([0.0, 0.0]) == [0.0, 0.0] and spark_heights([]) == []
def _target(reach="up", ok_minutes=1):
at = NOW - timedelta(minutes=ok_minutes)
return {"id": "t1", "name": T, "reach": reach, "says": "", "ok_at": at, "scraped_at": NOW}
async def _page(app, user=True, **ctx):
from quart import session
async with app.test_request_context("/plugins/prometheus/widget"):
if user:
session["user_id"] = "u1"
session["user_role"] = "viewer"
ctx.setdefault("spark_hours", 24)
page = await render_plugin_template(app, "prometheus", "prometheus/widget.html", **ctx)
return " ".join(page.split())
async def test_the_widget_leads_with_the_verdict_then_problems_first(app):
view = render(check(TEMPLATE), T, latest(SERIES), NOW,
history={(T, "fa_disk_free_bytes"): [50.0, 100.0]})
page = await _page(app, problem="", view=view, target=_target())
assert page.index("3 need a look. Backup targets: Photos is late.") < page.index("Backup targets</div>")
assert 'class="w-verdict tone-warn"' in page
# Photos (late) before Steward (fresh) before Old (disabled, quiet).
targets = page[page.index("Backup targets</div>"):page.index("Storage</div>")]
assert targets.index("Photos") < targets.index("Steward") < targets.index("Old")
assert '<span class="w-sub tone-warn">is late</span>' in page
assert "<th>Now</th>" in page and '<th class="w-wide-only">Last backup</th>' in page
assert "16.2 TiB of 41.8 TiB" in page and 'class="meter-fill meter-ok" style="width:38.7%"' in page
assert '<span class="w-bars w-wide-only"' in page and 'style="height:50.0%"' in page
assert "<th>files</th><th>trash</th>" in page and "437.7 GiB" in page
assert "version 2026.10.12.1717 · commit 3605d0a" in page
assert '<a href="/plugins/prometheus/targets/t1">Archives</a>' in page
assert not hex_colours(page), hex_colours(page)
async def test_the_healthy_rest_folds_after_six(app):
many = {**SERIES}
for i in range(9):
many.update(states(f"target=T{i},target_id=x{i}", "fresh", ["fresh", "late"]))
view = render(check(TEMPLATE), T, latest(many), NOW)
page = await _page(app, problem="", view=view, target=_target())
# 1 late, 10 fresh, 1 quiet in Backup targets: the late one, six of the eleven others.
assert "+5 more healthy" in page
async def test_a_target_not_answering_says_so_first(app):
view = render(check(TEMPLATE), T, latest(SERIES), NOW)
page = await _page(app, problem="", view=view, target=_target("down", ok_minutes=90))
assert '<div class="w-verdict tone-crit">Archives isn\'t answering; these values are from' in page
assert page.index("isn't answering") < page.index("3 need a look")
async def test_a_widget_that_cant_draw_says_why(app):
page = await _page(app, problem="The target this widget showed has been deleted.", view=None, target=None)
assert '<div class="w-verdict tone-warn">The target this widget showed has been deleted.</div>' in page
async def test_a_shared_dashboard_gets_no_links(app):
view = render(check(TEMPLATE), T, latest(SERIES), NOW)
page = await _page(app, user=False, problem="", view=view, target=_target())
assert "/plugins/prometheus/targets/t1" not in page and "Archives, scraped" in page
def test_the_widget_names_itself_after_its_template():
assert asyncio.run(library.default_title(None, {"template": "bundled:fabled_archives"})) == "Fabled Archives"
assert asyncio.run(library.default_title(None, {"template": "bundled:nope"})) is None