autonomic-bot
9a7772563a
Push builds have been RED on the lint step since ~build 209 from accumulated formatting drift. This is the mechanical cleanup: ruff format + ruff --fix (UP038 isinstance unions, SIM105 contextlib.suppress, UP031 f-strings, SIM115 tempfile context manager), shfmt -i 2 -ci, nixpkgs-fmt/statix/deadnix (merged attrsets, dropped unused lib args), yamllint, and shell quoting fixes in tests/lasuite-docs/setup_custom_tests.sh. No behaviour changes intended; lint: PASS, unit tests: 138 passed.
81 lines
3.8 KiB
Python
81 lines
3.8 KiB
Python
"""Unit tests for the F2-12 owned upgrade-convergence wait (P7-negative / non-vacuousness).
|
|
|
|
F2-12 fix: the upgrade chaos redeploy runs with `abra … -c` (skips abra's own convergence monitor),
|
|
and the harness OWNS the verification via `lifecycle.wait_healthy` (services N/N + app HEALTH_PATH)
|
|
then `lifecycle.wait_ready_probes` (recipe READY_PROBE, e.g. collabora WOPI discovery → 200). Skipping
|
|
abra's monitor is only acceptable if the replacement genuinely FAILS a broken convergence rather than
|
|
green-washing it (plan §7.1 / Adversary pre-claim recon 2026-05-29). These tests prove exactly that,
|
|
deterministically (fake clock, no deploy):
|
|
- wait_ready_probes RAISES when the probe never returns an OK status (stuck service).
|
|
- wait_ready_probes RETURNS when the probe reaches OK (and is a no-op without a READY_PROBE).
|
|
- wait_healthy RAISES when services never converge, and when they converge but HTTP never serves OK.
|
|
So `-c` + owned-wait is non-vacuous: a genuinely-broken upgrade stays RED.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import os
|
|
import sys
|
|
|
|
import pytest
|
|
|
|
sys.path.insert(0, os.path.join(os.path.dirname(__file__), "..", "..", "runner"))
|
|
from harness import lifecycle as lc # noqa: E402
|
|
|
|
|
|
def _fake_clock(monkeypatch):
|
|
"""Install a fake monotonic clock on lifecycle.time so sleeps advance it without real waiting."""
|
|
state = {"t": 1000.0}
|
|
monkeypatch.setattr(lc.time, "time", lambda: state["t"])
|
|
monkeypatch.setattr(lc.time, "sleep", lambda s: state.__setitem__("t", state["t"] + s))
|
|
return state
|
|
|
|
|
|
_DRIVE_META = {
|
|
"READY_PROBE": lambda d: [
|
|
{"host": f"collabora-{d}", "path": "/hosting/discovery", "ok": (200,)}
|
|
]
|
|
}
|
|
|
|
|
|
def test_wait_ready_probes_raises_when_never_ready(monkeypatch):
|
|
"""A READY_PROBE that never returns 200 (collabora discovery wedged) must raise TimeoutError —
|
|
NOT silently pass. This is what makes `-c` + owned-wait catch a genuinely stuck upgrade."""
|
|
_fake_clock(monkeypatch)
|
|
monkeypatch.setattr(lc, "http_get", lambda host, path="/", timeout=15: 404)
|
|
with pytest.raises(TimeoutError):
|
|
lc.wait_ready_probes(_DRIVE_META, "lasu-x.ci.commoninternet.net", timeout=120)
|
|
|
|
|
|
def test_wait_ready_probes_returns_when_ready(monkeypatch):
|
|
"""When the probe reaches 200, it returns (no raise)."""
|
|
_fake_clock(monkeypatch)
|
|
monkeypatch.setattr(lc, "http_get", lambda host, path="/", timeout=15: 200)
|
|
lc.wait_ready_probes(_DRIVE_META, "lasu-x.ci.commoninternet.net", timeout=120) # no raise
|
|
|
|
|
|
def test_wait_ready_probes_noop_without_probe(monkeypatch):
|
|
"""A recipe with no READY_PROBE is a clean no-op (default behavior preserved for all recipes)."""
|
|
monkeypatch.setattr(lc, "http_get", lambda *a, **k: 599) # would fail if it were consulted
|
|
lc.wait_ready_probes({}, "x.ci.commoninternet.net", timeout=1) # no raise, no call
|
|
|
|
|
|
def test_wait_healthy_raises_when_services_never_converge(monkeypatch):
|
|
"""If swarm services never reach N/N, wait_healthy must raise (the upgrade op then fails the tier)."""
|
|
_fake_clock(monkeypatch)
|
|
monkeypatch.setattr(lc, "services_converged", lambda domain: False)
|
|
monkeypatch.setattr(lc, "http_get", lambda *a, **k: 200) # irrelevant; convergence gates first
|
|
with pytest.raises(TimeoutError):
|
|
lc.wait_healthy("x.ci.commoninternet.net", deploy_timeout=60, http_timeout=60)
|
|
|
|
|
|
def test_wait_healthy_raises_when_converged_but_never_serves(monkeypatch):
|
|
"""Services converged but the app never returns an OK status → wait_healthy raises (non-vacuous)."""
|
|
_fake_clock(monkeypatch)
|
|
monkeypatch.setattr(lc, "services_converged", lambda domain: True)
|
|
monkeypatch.setattr(lc, "http_get", lambda *a, **k: 502)
|
|
with pytest.raises(TimeoutError):
|
|
lc.wait_healthy(
|
|
"x.ci.commoninternet.net", ok_codes=(200,), deploy_timeout=60, http_timeout=60
|
|
)
|