Files
FabledSteward/tests/integration/test_docker.py
T
bvandeusen 578cc33cc0
CI / lint (push) Successful in 2s
CI / unit (push) Successful in 48s
CI / integration (push) Successful in 2m22s
CI / publish (push) Successful in 1m8s
feat(docker): ingest swarm topology + lifecycle events + health/restart alerts
Wires the agent's enriched + swarm payloads through the docker.persist_host_
samples capability:

  * Swarm topology — persist sample["swarm"] into docker_swarm_services /
    docker_swarm_nodes (upsert + prune stale, host-scoped so two managers don't
    clobber). Migration docker_005 adds services.placement_json for the
    task→node placement the agent now reports.
  * Lifecycle events — _derive_events (pure, unit-tested) diffs the newest
    snapshot against stored per-container state: start / stop / die (non-zero
    exit) / oom / health_change → docker_events rows. Skipped on a host's first
    snapshot so the baseline doesn't emit a start per existing container.
  * Alerts — record restart_count (always) and is_healthy (1.0/0.0, only when a
    HEALTHCHECK exists) alongside cpu/mem, under host-scoped resource names;
    METRIC_CATALOG[docker] gains restart_count + is_healthy so they're alertable.

host_agent ingest captures the newest sample's swarm object and threads it to
the capability (now persist_host_docker(session, host, snapshots, swarm=None));
invoked when containers OR swarm are present, under the same SAVEPOINT. Unit
tests cover the event-diff matrix; integration tests cover event derivation
across two snapshots and swarm topology round-trip (incl. placement).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC
2026-06-18 21:07:40 -04:00

222 lines
9.6 KiB
Python

"""Integration: per-host Docker schema + host-scoped persistence.
Validates the docker_002 migration (host_id columns + composite key) applied and
that persist_host_docker scopes containers/metrics by host so same-named
containers on different hosts no longer collide. Requires STEWARD_DATABASE_URL.
"""
from __future__ import annotations
import asyncio
import os
import uuid
from datetime import datetime, timezone
import pytest
pytestmark = pytest.mark.integration
_NEEDS_DB = pytest.mark.skipif(
not os.environ.get("STEWARD_DATABASE_URL"),
reason="integration test needs a live Postgres (STEWARD_DATABASE_URL)",
)
@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)
@_NEEDS_DB
def test_docker_tables_are_host_scoped(app):
from sqlalchemy import text
async def _go():
async with app.db_sessionmaker() as s:
# Both tables carry a host_id column after docker_002.
for tbl in ("docker_containers", "docker_metrics"):
cols = {r[0] for r in (await s.execute(text(
"SELECT column_name FROM information_schema.columns "
"WHERE table_name = :t"), {"t": tbl})).all()}
assert "host_id" in cols, f"{tbl} missing host_id"
# docker_containers is keyed by (host_id, name), not name alone.
pk_cols = {r[0] for r in (await s.execute(text(
"SELECT a.attname FROM pg_index i "
"JOIN pg_attribute a ON a.attrelid = i.indrelid "
" AND a.attnum = ANY(i.indkey) "
"WHERE i.indrelid = 'docker_containers'::regclass AND i.indisprimary"
))).all()}
assert pk_cols == {"host_id", "name"}
asyncio.run(_go())
@_NEEDS_DB
def test_events_and_swarm_tables_exist(app):
"""docker_004 created the lifecycle + swarm topology tables with host_id FKs."""
from sqlalchemy import text
async def _go():
async with app.db_sessionmaker() as s:
for tbl in ("docker_events", "docker_swarm_services", "docker_swarm_nodes"):
cols = {r[0] for r in (await s.execute(text(
"SELECT column_name FROM information_schema.columns "
"WHERE table_name = :t"), {"t": tbl})).all()}
assert cols, f"{tbl} missing entirely"
assert "host_id" in cols, f"{tbl} missing host_id"
# docker_events carries the lifecycle columns the ingest derivation writes.
ev_cols = {r[0] for r in (await s.execute(text(
"SELECT column_name FROM information_schema.columns "
"WHERE table_name = 'docker_events'"))).all()}
assert {"event", "container_name", "at", "detail"} <= ev_cols
asyncio.run(_go())
def _persist_fn(app):
"""Resolve persist_host_docker via the registered capability if the docker
plugin is loaded, else import it directly (the import is safe only when the
plugin is NOT loaded — otherwise its models are already mapped)."""
from steward.core.capabilities import has_capability, get_capability
if has_capability("docker.persist_host_samples"):
return get_capability("docker.persist_host_samples").fn
from plugins.docker.ingest import persist_host_docker
return persist_host_docker
@_NEEDS_DB
def test_persist_scopes_containers_by_host(app):
from sqlalchemy import text
from steward.models.hosts import Host
persist = _persist_fn(app)
now = datetime.now(timezone.utc)
snapshot = [(now, [{
"name": "web", "container_id": "abc123", "image": "nginx",
"status": "running", "cpu_pct": 12.5, "mem_pct": 30.0,
"mem_usage_bytes": 100, "mem_limit_bytes": 200, "ports": [], "started_at": None,
# enrichment (agent ≥ 1.4.0)
"health": "healthy", "restart_count": 2, "exit_code": 0, "oom_killed": False,
"compose_project": "stack", "service_name": "web",
"net_rx_bytes": 1000, "net_tx_bytes": 2000,
"blk_read_bytes": 4096, "blk_write_bytes": 8192,
}])]
async def _go():
async with app.db_sessionmaker() as s:
async with s.begin():
await s.execute(text("DELETE FROM docker_metrics"))
await s.execute(text("DELETE FROM docker_containers"))
await s.execute(text(
"DELETE FROM plugin_metrics WHERE source_module = 'docker'"))
h1 = Host(id=str(uuid.uuid4()), name="alpha", address="10.0.0.1")
h2 = Host(id=str(uuid.uuid4()), name="beta", address="10.0.0.2")
s.add_all([h1, h2])
await s.flush()
# Same container name on two hosts — must not collide now.
await persist(s, h1, snapshot)
await persist(s, h2, snapshot)
web_rows = (await s.execute(text(
"SELECT COUNT(*) FROM docker_containers WHERE name = 'web'"))).scalar()
metric_hosts = (await s.execute(text(
"SELECT COUNT(DISTINCT host_id) FROM docker_metrics "
"WHERE container_name = 'web'"))).scalar()
# Alert pipeline received host-scoped resource names.
resources = {r[0] for r in (await s.execute(text(
"SELECT DISTINCT resource_name FROM plugin_metrics "
"WHERE source_module = 'docker'"))).all()}
# Enrichment columns persisted (check one host's row).
enrich = (await s.execute(text(
"SELECT health, restart_count, service_name, net_rx_bytes "
"FROM docker_containers WHERE name = 'web' LIMIT 1"))).first()
return web_rows, metric_hosts, resources, enrich
web_rows, metric_hosts, resources, enrich = asyncio.run(_go())
assert web_rows == 2 # one 'web' per host, keyed by (host_id, name)
assert metric_hosts == 2 # time-series rows scoped per host
assert resources == {"alpha/web", "beta/web"}
assert enrich == ("healthy", 2, "web", 1000) # enrichment round-trips
@_NEEDS_DB
def test_lifecycle_events_derived_across_snapshots(app):
from sqlalchemy import text
from steward.models.hosts import Host
persist = _persist_fn(app)
t1 = datetime(2026, 6, 19, 12, 0, 0, tzinfo=timezone.utc)
t2 = datetime(2026, 6, 19, 12, 0, 30, tzinfo=timezone.utc)
running = {"name": "web", "container_id": "abc", "image": "nginx",
"status": "running", "cpu_pct": 5.0, "mem_pct": 10.0,
"restart_count": 0, "exit_code": None, "oom_killed": False,
"health": "healthy"}
died = {**running, "status": "exited", "cpu_pct": None,
"exit_code": 137, "oom_killed": True, "health": None}
async def _go():
async with app.db_sessionmaker() as s:
async with s.begin():
await s.execute(text("DELETE FROM docker_events"))
await s.execute(text("DELETE FROM docker_metrics"))
await s.execute(text("DELETE FROM docker_containers"))
h = Host(id=str(uuid.uuid4()), name="evt", address="10.9.9.9")
s.add(h)
await s.flush()
# Baseline snapshot establishes state — must NOT emit a start.
await persist(s, h, [(t1, [running])])
# Second snapshot: the container OOM-died.
await persist(s, h, [(t2, [died])])
hid = h.id
rows = (await s.execute(text(
"SELECT event, detail FROM docker_events WHERE host_id = :h "
"ORDER BY event"), {"h": hid})).all()
return [(r[0], r[1]) for r in rows]
events = asyncio.run(_go())
# baseline emitted nothing; the death emitted both die (exit 137) and oom.
assert ("die", "exit 137") in events
assert ("oom", None) in events
assert not any(e[0] == "start" for e in events)
@_NEEDS_DB
def test_swarm_topology_persisted(app):
from sqlalchemy import text
from steward.models.hosts import Host
persist = _persist_fn(app)
swarm = {
"services": [{"service_name": "web", "mode": "replicated",
"desired": 3, "running": 2, "image": "nginx",
"placement": [{"node_id": "n1", "running": 2}]}],
"nodes": [{"node_id": "n1", "hostname": "mgr", "role": "manager",
"availability": "active", "status": "ready", "leader": True}],
}
async def _go():
async with app.db_sessionmaker() as s:
async with s.begin():
await s.execute(text("DELETE FROM docker_swarm_services"))
await s.execute(text("DELETE FROM docker_swarm_nodes"))
h = Host(id=str(uuid.uuid4()), name="mgr", address="10.8.8.8")
s.add(h)
await s.flush()
# No containers in this sample — swarm must still persist.
await persist(s, h, [], swarm)
hid = h.id
svc = (await s.execute(text(
"SELECT mode, desired, running, image, placement_json "
"FROM docker_swarm_services WHERE host_id = :h"), {"h": hid})).first()
node = (await s.execute(text(
"SELECT role, availability, status, leader "
"FROM docker_swarm_nodes WHERE host_id = :h"), {"h": hid})).first()
return svc, node
svc, node = asyncio.run(_go())
assert svc[0] == "replicated" and svc[1] == 3 and svc[2] == 2 and svc[3] == "nginx"
assert "n1" in svc[4] # placement JSON carries the node id
assert node[0] == "manager" and node[2] == "ready" and node[3] is True