**Phase 09 — final verification pass (all 4 task files already in `complete/`; nothing re-implemented)**
- Verified `GET /api/disks/snapshot` live: 3 hosts / 18 whole-`/sys/block` rows, `smartctl_available` three-way, per-row `severity`, API-summed `fleet`; 404 `{"configured": false}` proven *without dialling* (spies on `httpx.AsyncClient.get` + `collectors.fetch_snapshot`).
- Verified `/api/disks/history` live (`disk.temp_c@dev`, unknown names → empty series) + `test_store.py`'s 300 s rollup/prune buckets; `/api/metrics?source=disks` answers the `disks` family.
- Browser against live data: 18 drive cards (never a partition/mount), collapsed per-host `Charts` with `aria-label`, 3 `.stale` grids + one `.banner.info` why-line and **0 `.fail` classes / green `disks` dot** → criteria 3 and 4 hold; new shots `data/logs/09-verify-disks-live-dimmed.png`, `09-verify-overview-live.png` (13 total for the phase).
- Real-hardware criterion: read 4 driveripper drives through `ops`' `sudo -n smartctl -i -A -H` and ran the live output through `host_agent.parse_smart_output` → `sources/disks.parse_disks_snapshot`: `sda` 35 °C / 30 423 h, `nvme1n1` 58 % wear, `nvme5n1` 13 uncorrectable → `alarm_disks: 1`, temps graded `warn`; consistent with the 16 committed fixtures.
- `bash .agents/validate.sh` → **EXIT=0**; `uv run pytest --cov` → **769 passed, 11 skipped, coverage 98.33%** (floor 94); `ruff check src tests agent` clean; `pyright` 0 errors; `node --check` clean; hygiene + dependency advisories clean. `uv run pytest -m live --live` → 10 passed, 1 skipped (agents still predate the SMART pass). Dev server healthy: `/api/health` `ok`, 9/9 sources.
- Docs fixed: AGENTS.md test count 698→769 and coverage 98.2→98.33 %; `disks` added to the event dashboard-name list and to the "every feed renders `Dash.eventText`" sentence; README `POLL_MS_*` row now states `/disks`' deliberate fixed 30 s; new Known gap 4 records the dev-DB `database is locked` writes (reported/recovered per invariant 4, ~590 MB DB / ~70 MB WAL, not reproduced by a controlled 25 s external read txn).
- Deviations: **no live-infra change** — granting the agent user the `disk` group and redeploying `host_agent.py` on driveripper/deskwork stay the documented paired operator action (redeploying without the grant yields 16 permission-denied lines; a sudo-spawning agent would break the confirmed "direct read, no sudo" assumption). Store-lock issue left to its own phase rather than patched speculatively.
- Next pending phase: `10_dev_direct_dial`.
219 lines
9.3 KiB
Python
219 lines
9.3 KiB
Python
"""Configuration: env-only, no defaults that point at somebody else's machine."""
|
|
from __future__ import annotations
|
|
|
|
import importlib
|
|
import os
|
|
from dataclasses import FrozenInstanceError
|
|
from typing import Any
|
|
|
|
import pytest
|
|
|
|
from dashboards import config
|
|
from dashboards.config import Settings, parse_targets
|
|
|
|
|
|
def test_parses_whitespace_and_comma_separated_targets() -> None:
|
|
targets, problems = parse_targets("a=http://h1:9120 b=http://h2:9120,c=http://h3:9120")
|
|
assert [t.name for t in targets] == ["a", "b", "c"]
|
|
assert [t.url for t in targets] == ["http://h1:9120", "http://h2:9120", "http://h3:9120"]
|
|
assert problems == []
|
|
|
|
|
|
def test_trailing_slash_is_stripped_so_urls_do_not_double_up() -> None:
|
|
targets, _ = parse_targets("a=http://h1:9120/")
|
|
assert targets[0].url == "http://h1:9120"
|
|
|
|
|
|
def test_one_bad_entry_does_not_sink_the_others() -> None:
|
|
targets, problems = parse_targets("good=http://h:1 broken-no-equals another=ftp://h:1")
|
|
assert [t.name for t in targets] == ["good"]
|
|
assert len(problems) == 2
|
|
assert any("not name=url" in p for p in problems)
|
|
assert any("http://" in p for p in problems)
|
|
|
|
|
|
def test_empty_input_yields_no_targets_and_no_noise() -> None:
|
|
assert parse_targets("") == ([], [])
|
|
assert parse_targets(" \n ") == ([], [])
|
|
|
|
|
|
def test_token_in_url_becomes_a_bearer_header_and_stays_out_of_the_display_url() -> None:
|
|
targets, problems = parse_targets("driveripper=http://SEKRIT@driveripper:9120")
|
|
assert problems == []
|
|
target = targets[0]
|
|
assert target.headers == {"Authorization": "Bearer SEKRIT"}
|
|
assert target.url == "http://driveripper:9120", "the token must not stay in the URL"
|
|
assert target.display_url == "http://driveripper:9120"
|
|
assert "SEKRIT" not in target.display_url
|
|
|
|
|
|
@pytest.mark.parametrize("raw,shown", [
|
|
("http://SEKRIT@host:9120", "http://host:9120"),
|
|
("https://user:pass@host:8000/metrics", "https://host:8000/metrics"),
|
|
("HTTP://SEKRIT@host:1", "HTTP://host:1"),
|
|
("http://host:9120", "http://host:9120"), # clean URLs pass through untouched
|
|
("", ""),
|
|
])
|
|
def test_strip_userinfo_is_the_one_redaction_every_url_leaves_by(raw: str, shown: str) -> None:
|
|
"""`strip_userinfo` is what `/api/config` and every `UpstreamError` message use, so it has
|
|
to survive the shapes a URL actually comes in: both schemes, `user:pass@`, a path, and the
|
|
common no-credentials case where it must change nothing (invariant 6).
|
|
"""
|
|
assert config.strip_userinfo(raw) == shown
|
|
|
|
|
|
def test_target_names_are_restricted_because_they_are_db_keys() -> None:
|
|
"""Names key the history DB and appear in URLs, so they are validated, not guessed.
|
|
|
|
Whitespace separates entries, which means "Bad Name=http://h" is two entries — both
|
|
reported, because the operator needs to see exactly what was rejected.
|
|
"""
|
|
targets, problems = parse_targets("Bad=http://h:1 /=http://h:1 ok-1_2=http://h:1")
|
|
assert [t.name for t in targets] == ["ok-1_2"]
|
|
assert len(problems) == 2
|
|
assert any("Bad" in p for p in problems)
|
|
|
|
_targets, problems = parse_targets("Bad Name=http://h:1")
|
|
assert len(problems) == 2, "a space splits the entry, so both halves are rejected"
|
|
|
|
|
|
def test_settings_default_to_no_sources_configured() -> None:
|
|
settings = Settings()
|
|
assert settings.host_targets == ()
|
|
assert settings.gufo_url == "" and settings.vllm_url == ""
|
|
assert settings.nut_host == ""
|
|
assert settings.nut_enabled is False
|
|
assert settings.any_source_enabled is False
|
|
|
|
|
|
def test_the_disk_source_defaults_to_a_five_minute_cadence() -> None:
|
|
"""SMART is not a per-second signal (phase 09), and the fetch waits for the agent's
|
|
own 20 s smartctl budget. Both are env-only, like every other interval."""
|
|
settings = Settings()
|
|
assert settings.disks_interval == 300.0
|
|
assert settings.disks_timeout == 30.0
|
|
|
|
|
|
def test_the_disk_cadence_is_an_env_var_like_every_other_one(monkeypatch: pytest.MonkeyPatch) -> None:
|
|
"""`DISKS_INTERVAL_SECONDS=60` is how an operator watches a failing drive closely."""
|
|
monkeypatch.setenv("DISKS_INTERVAL_SECONDS", "60")
|
|
monkeypatch.setenv("DISKS_TIMEOUT_SECONDS", "45")
|
|
reloaded = importlib.reload(config)
|
|
try:
|
|
assert reloaded.SETTINGS.disks_interval == 60.0
|
|
assert reloaded.SETTINGS.disks_timeout == 45.0
|
|
finally:
|
|
monkeypatch.delenv("DISKS_INTERVAL_SECONDS")
|
|
monkeypatch.delenv("DISKS_TIMEOUT_SECONDS")
|
|
importlib.reload(config)
|
|
|
|
|
|
def test_actions_need_the_flag_and_are_dead_under_read_only() -> None:
|
|
assert Settings(nut_allow_actions=True, read_only=False).actions_allowed is True
|
|
assert Settings(nut_allow_actions=True, read_only=True).actions_allowed is False
|
|
assert Settings(nut_allow_actions=False, read_only=False).actions_allowed is False
|
|
|
|
|
|
def test_nut_host_accepts_a_nut_scheme_prefix() -> None:
|
|
assert Settings(nut_host="nut://ups-box:3493").nut_enabled is True
|
|
|
|
|
|
def test_load_reads_host_targets_from_the_environment(monkeypatch: pytest.MonkeyPatch) -> None:
|
|
monkeypatch.setenv("HOST_TARGETS", "deskwork=http://127.0.0.1:9122")
|
|
settings = config.load()
|
|
assert [t.name for t in settings.host_targets] == ["deskwork"]
|
|
|
|
|
|
def test_env_driven_scalars_are_read_at_import(monkeypatch: pytest.MonkeyPatch) -> None:
|
|
"""The dataclass defaults read os.environ, so a reload must pick up a changed env."""
|
|
monkeypatch.setenv("DASH_PORT", "8123")
|
|
monkeypatch.setenv("RETAIN_RAW_HOURS", "6")
|
|
monkeypatch.setenv("NUT_ALLOW_ACTIONS", "1")
|
|
reloaded = importlib.reload(config)
|
|
try:
|
|
assert reloaded.SETTINGS.port == 8123
|
|
assert reloaded.SETTINGS.raw_keep_hours == 6
|
|
assert reloaded.SETTINGS.actions_allowed is True
|
|
finally:
|
|
monkeypatch.delenv("DASH_PORT")
|
|
monkeypatch.delenv("RETAIN_RAW_HOURS")
|
|
monkeypatch.delenv("NUT_ALLOW_ACTIONS")
|
|
importlib.reload(config)
|
|
|
|
|
|
def test_flags_accept_the_usual_spellings(monkeypatch: pytest.MonkeyPatch) -> None:
|
|
for truthy in ("1", "true", "TRUE", "yes", "on"):
|
|
monkeypatch.setenv("DASH_READ_ONLY", truthy)
|
|
assert config._flag("DASH_READ_ONLY") is True, truthy
|
|
for falsey in ("0", "false", "", "no"):
|
|
monkeypatch.setenv("DASH_READ_ONLY", falsey)
|
|
assert config._flag("DASH_READ_ONLY") is False, falsey
|
|
|
|
|
|
def test_numeric_env_must_be_a_number_or_the_default_wins(monkeypatch: pytest.MonkeyPatch) -> None:
|
|
monkeypatch.setenv("HOST_INTERVAL_SECONDS", "not-a-number")
|
|
assert config._num("HOST_INTERVAL_SECONDS", 10.0) == 10.0
|
|
monkeypatch.setenv("HOST_INTERVAL_SECONDS", "2.5")
|
|
assert config._num("HOST_INTERVAL_SECONDS", 10.0) == 2.5
|
|
|
|
|
|
def test_settings_is_immutable_so_nobody_mutates_shared_state() -> None:
|
|
settings: Any = Settings()
|
|
with pytest.raises(FrozenInstanceError):
|
|
settings.port = 1 # type: ignore[misc]
|
|
|
|
|
|
# ---- the .env convenience -------------------------------------------------- #
|
|
|
|
def test_no_env_file_is_the_normal_case(monkeypatch: pytest.MonkeyPatch, tmp_path: Any) -> None:
|
|
monkeypatch.setenv("DASH_ENV_FILE", str(tmp_path / "absent.env"))
|
|
assert config.load_env_file() is None
|
|
|
|
|
|
def test_env_file_seeds_values_a_deployed_env_does_not_set(monkeypatch: pytest.MonkeyPatch,
|
|
tmp_path: Any) -> None:
|
|
env = tmp_path / ".env"
|
|
env.write_text("DASH_PORT=9999\nNUT_HOST=ups.internal\n")
|
|
monkeypatch.setenv("DASH_ENV_FILE", str(env))
|
|
monkeypatch.delenv("DASH_PORT", raising=False)
|
|
monkeypatch.delenv("NUT_HOST", raising=False)
|
|
assert config.load_env_file() == env
|
|
# Settings reads its defaults when the module is imported, so the reload is what a real
|
|
# process gets for free and what the test has to ask for explicitly
|
|
reloaded = importlib.reload(config)
|
|
assert reloaded.SETTINGS.port == 9999
|
|
assert reloaded.SETTINGS.nut_host == "ups.internal"
|
|
|
|
|
|
def test_a_real_environment_variable_beats_the_file(monkeypatch: pytest.MonkeyPatch,
|
|
tmp_path: Any) -> None:
|
|
"""The container/quadlet env is the source of truth; the file is for laptops."""
|
|
env = tmp_path / ".env"
|
|
env.write_text("DASH_PORT=9999\n")
|
|
monkeypatch.setenv("DASH_ENV_FILE", str(env))
|
|
monkeypatch.setenv("DASH_PORT", "7777")
|
|
reloaded = importlib.reload(config)
|
|
assert os.environ["DASH_PORT"] == "7777"
|
|
assert reloaded.SETTINGS.port == 7777
|
|
|
|
|
|
def test_env_file_expands_other_variables(monkeypatch: pytest.MonkeyPatch, tmp_path: Any) -> None:
|
|
"""`.env.example` shows `DASH_DB=${HOME}/...` — the file has to mean what it shows."""
|
|
env = tmp_path / ".env"
|
|
env.write_text("MY_BASE=/srv/dash\nDASH_DB=${MY_BASE}/dashboards.sqlite3\n")
|
|
monkeypatch.setenv("DASH_ENV_FILE", str(env))
|
|
monkeypatch.delenv("MY_BASE", raising=False)
|
|
monkeypatch.delenv("DASH_DB", raising=False)
|
|
reloaded = importlib.reload(config)
|
|
assert reloaded.SETTINGS.db_path == "/srv/dash/dashboards.sqlite3"
|
|
|
|
|
|
def test_the_example_file_is_never_loaded(monkeypatch: pytest.MonkeyPatch, tmp_path: Any) -> None:
|
|
example = tmp_path / ".env.example"
|
|
example.write_text("NUT_HOST=should-not-load\n")
|
|
monkeypatch.setenv("DASH_ENV_FILE", str(example))
|
|
monkeypatch.delenv("NUT_HOST", raising=False)
|
|
reloaded = importlib.reload(config)
|
|
assert reloaded.load_env_file() is None
|
|
assert reloaded.SETTINGS.nut_host == ""
|