Skip to content

fix: permit transactional maintenance health #58

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Nickfost merged 1 commit into from
Jul 28, 2026
Merged

fix: permit transactional maintenance health #58

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion scripts/health.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -367,8 +367,9 @@ def collect_snapshot(values: dict[str, str], *, root: Path = Path("/"), run: Run
timers["updates"] = _unit_state(run, "apt-daily-upgrade.timer", timer=True, max_age_seconds=172800)
services["updates"] = _unit_state(run, "apt-daily-upgrade.service")
if values.get("CI_FLEET_HEALTH_BOOTSTRAP") == "1":
# ponytail: activation validates unit installation separately; scheduled runs verify live timers after enablement.
# ponytail: activation validates unit installation separately; scheduled runs verify maintenance state after commit.
timers = {name: "ok" for name in timers}
services = {name: "ok" for name in services}
instance = values.get("CI_FLEET_INSTANCE", "unknown")
stale = _stale_resources(run, instance) if docker_ok else {"containers": 0, "networks": 0, "volumes": 0}
stale["images"] = _count(run, ["docker", "images", "-q", "--filter", "dangling=true", "--filter", "label=io.randomdevelopment.ci-fleet.managed=true"]) if docker_ok else 0
Expand Down
2 changes: 1 addition & 1 deletion scripts/install-worker-controller.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -929,7 +929,7 @@ restore_checkpoint() {
failed=1
else
compose "$release_dir" "$rendered_env" up -d --no-deps controller || failed=1
if ((failed == 0)); then run_health_check "$release_dir" "$rendered_env" || failed=1; fi
if ((failed == 0)); then run_health_check "$release_dir" "$rendered_env" true || failed=1; fi
fi
fi
fi
Expand Down
3 changes: 2 additions & 1 deletion scripts/test_health.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -157,7 +157,7 @@ def test_collector_uses_sustained_metrics_and_all_service_units(self) -> None:
pressure = root / "proc/pressure/memory"
pressure.write_text("some avg10=0.00 avg60=0.05 avg300=0.20 total=1\n")
def run(args):
output = "yes\n" if args[0] == "timedatectl" else "success\n" if args[0] == "systemctl" else ""
output = "yes\n" if args[0] == "timedatectl" else "failed\n" if args[:2] == ["systemctl", "show"] and args[2].endswith(".service") else "success\n" if args[0] == "systemctl" else ""
return health.subprocess.CompletedProcess(args, 1 if args[:2] == ["docker", "info"] else 0, output, "")

original_load, original_cpus = health.os.getloadavg, health.os.cpu_count
Expand All @@ -183,6 +183,7 @@ def run(args):
run=run,
)
self.assertEqual(set(remote["services"]), {"cleanup", "drift", "reconcile"})
self.assertEqual(set(remote["services"].values()), {"ok"})
self.assertEqual(set(remote["timers"]), {"health", "cleanup", "drift", "reconcile"})
self.assertEqual(remote["reconciliation"]["status"], "bootstrap")
(root / "etc").mkdir()
Expand Down