Agents already exposed POST /teardown; the master was missing the plumbing
to reach it. Add:
- POST /api/v1/swarm/hosts/{uuid}/teardown — admin-gated. Body
{decky_id: str|null}: null tears the whole host, a value tears one decky.
On worker failure the master returns 502 and leaves DB shards intact so
master and agent stay aligned.
- BaseRepository.delete_decky_shard(name) + sqlmodel impl for per-decky
cleanup after a single-decky teardown.
- SwarmHosts page: "Teardown all" button (keeps host enrolled).
- SwarmDeckies page: per-row "Teardown" button.
Also exclude setuptools' build/ staging dir from the enrollment tarball —
`pip install -e` on the master generates build/lib/decnet_web/node_modules
and the bundle walker was leaking it to agents. Align pyproject's bandit
exclude with the git-hook invocation so both skip decnet/templates/.
166 lines
5.1 KiB
Python
166 lines
5.1 KiB
Python
"""POST /swarm/hosts/{uuid}/teardown — per-host and per-decky remote teardown."""
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
from datetime import datetime, timezone
|
|
from typing import Optional
|
|
|
|
import pytest
|
|
|
|
from decnet.web.router.swarm_mgmt import api_teardown_host as mod
|
|
|
|
|
|
class _FakeAgent:
|
|
def __init__(self, *a, **kw):
|
|
_FakeAgent.calls.append(("init", kw.get("host", a[0] if a else None)))
|
|
self._host = kw.get("host", a[0] if a else None)
|
|
|
|
async def __aenter__(self):
|
|
return self
|
|
|
|
async def __aexit__(self, *exc):
|
|
return None
|
|
|
|
async def teardown(self, decky_id: Optional[str] = None) -> dict:
|
|
_FakeAgent.calls.append(("teardown", decky_id))
|
|
return {"status": "torn_down", "decky_id": decky_id}
|
|
|
|
|
|
class _FailingAgent(_FakeAgent):
|
|
async def teardown(self, decky_id: Optional[str] = None) -> dict:
|
|
raise RuntimeError("network unreachable")
|
|
|
|
|
|
@pytest.fixture
|
|
def fake_agent(monkeypatch):
|
|
_FakeAgent.calls = []
|
|
monkeypatch.setattr(mod, "AgentClient", _FakeAgent)
|
|
return _FakeAgent
|
|
|
|
|
|
@pytest.fixture
|
|
def failing_agent(monkeypatch):
|
|
_FailingAgent.calls = []
|
|
monkeypatch.setattr(mod, "AgentClient", _FailingAgent)
|
|
return _FailingAgent
|
|
|
|
|
|
async def _seed_host(repo, *, name="worker-a", uuid="h-1") -> str:
|
|
await repo.add_swarm_host({
|
|
"uuid": uuid,
|
|
"name": name,
|
|
"address": "10.0.0.9",
|
|
"agent_port": 8765,
|
|
"status": "active",
|
|
"client_cert_fingerprint": "f" * 64,
|
|
"cert_bundle_path": "",
|
|
"use_ipvlan": False,
|
|
"enrolled_at": datetime.now(timezone.utc),
|
|
"last_heartbeat": None,
|
|
})
|
|
return uuid
|
|
|
|
|
|
async def _seed_shard(repo, *, host_uuid: str, decky_name: str) -> None:
|
|
await repo.upsert_decky_shard({
|
|
"decky_name": decky_name,
|
|
"host_uuid": host_uuid,
|
|
"services": json.dumps(["ssh"]),
|
|
"state": "running",
|
|
"last_error": None,
|
|
"updated_at": datetime.now(timezone.utc),
|
|
})
|
|
|
|
|
|
@pytest.mark.anyio
|
|
async def test_teardown_all_deckies_on_host(client, auth_token, fake_agent):
|
|
from decnet.web.dependencies import repo
|
|
uuid = await _seed_host(repo, name="tear-all", uuid="tear-all-uuid")
|
|
await _seed_shard(repo, host_uuid=uuid, decky_name="decky1")
|
|
await _seed_shard(repo, host_uuid=uuid, decky_name="decky2")
|
|
|
|
resp = await client.post(
|
|
f"/api/v1/swarm/hosts/{uuid}/teardown",
|
|
headers={"Authorization": f"Bearer {auth_token}"},
|
|
json={},
|
|
)
|
|
assert resp.status_code == 200, resp.text
|
|
body = resp.json()
|
|
assert body["ok"] is True
|
|
assert body["decky_id"] is None
|
|
|
|
assert ("teardown", None) in fake_agent.calls
|
|
remaining = await repo.list_decky_shards(uuid)
|
|
assert remaining == []
|
|
|
|
|
|
@pytest.mark.anyio
|
|
async def test_teardown_single_decky(client, auth_token, fake_agent):
|
|
from decnet.web.dependencies import repo
|
|
uuid = await _seed_host(repo, name="tear-one", uuid="tear-one-uuid")
|
|
await _seed_shard(repo, host_uuid=uuid, decky_name="decky-keep")
|
|
await _seed_shard(repo, host_uuid=uuid, decky_name="decky-drop")
|
|
|
|
resp = await client.post(
|
|
f"/api/v1/swarm/hosts/{uuid}/teardown",
|
|
headers={"Authorization": f"Bearer {auth_token}"},
|
|
json={"decky_id": "decky-drop"},
|
|
)
|
|
assert resp.status_code == 200, resp.text
|
|
body = resp.json()
|
|
assert body["decky_id"] == "decky-drop"
|
|
|
|
assert ("teardown", "decky-drop") in fake_agent.calls
|
|
remaining = {s["decky_name"] for s in await repo.list_decky_shards(uuid)}
|
|
assert remaining == {"decky-keep"}
|
|
|
|
|
|
@pytest.mark.anyio
|
|
async def test_teardown_unknown_host_404(client, auth_token, fake_agent):
|
|
resp = await client.post(
|
|
"/api/v1/swarm/hosts/does-not-exist/teardown",
|
|
headers={"Authorization": f"Bearer {auth_token}"},
|
|
json={},
|
|
)
|
|
assert resp.status_code == 404
|
|
|
|
|
|
@pytest.mark.anyio
|
|
async def test_teardown_agent_failure_502(client, auth_token, failing_agent):
|
|
"""When the worker is unreachable the DB shards MUST NOT be deleted —
|
|
otherwise the master's view diverges from reality."""
|
|
from decnet.web.dependencies import repo
|
|
uuid = await _seed_host(repo, name="tear-fail", uuid="tear-fail-uuid")
|
|
await _seed_shard(repo, host_uuid=uuid, decky_name="survivor")
|
|
|
|
resp = await client.post(
|
|
f"/api/v1/swarm/hosts/{uuid}/teardown",
|
|
headers={"Authorization": f"Bearer {auth_token}"},
|
|
json={},
|
|
)
|
|
assert resp.status_code == 502
|
|
|
|
remaining = {s["decky_name"] for s in await repo.list_decky_shards(uuid)}
|
|
assert remaining == {"survivor"}
|
|
|
|
|
|
@pytest.mark.anyio
|
|
async def test_teardown_non_admin_forbidden(client, viewer_token, fake_agent):
|
|
from decnet.web.dependencies import repo
|
|
uuid = await _seed_host(repo, name="tear-guard", uuid="tear-guard-uuid")
|
|
resp = await client.post(
|
|
f"/api/v1/swarm/hosts/{uuid}/teardown",
|
|
headers={"Authorization": f"Bearer {viewer_token}"},
|
|
json={},
|
|
)
|
|
assert resp.status_code == 403
|
|
|
|
|
|
@pytest.mark.anyio
|
|
async def test_teardown_no_auth_401(client, fake_agent):
|
|
resp = await client.post(
|
|
"/api/v1/swarm/hosts/whatever/teardown",
|
|
json={},
|
|
)
|
|
assert resp.status_code == 401
|