Robert Wojciechowski
6b94242110
* feat: add Runpod serverless image collector pools * fix: prevent worker settings clipping and remove preview copy * fix: initialize worker settings before mounting the form * fix: validate complete serverless requests against 10 MB limit * fix: simplify serverless cards and keep status dots circular * fix: match serverless endpoint box to master card * fix: preserve serverless capacity and recovery across job failures Retain durable reservations until provider termination or explicit recovery, protect unresolved pools and credentials, and add scoped reconciliation. Consolidate lifecycle and secret policies and split prompt/media stages with regression coverage. * fix(serverless): disable Clear saved API key while the worker is enabled The checkbox read "Clear saved API key (disable pool first)", but the panel never defines "pool" and saving with it ticked on an enabled worker was rejected by the backend. - Disable the checkbox while the worker is enabled, with a tooltip saying to turn the worker off first. Also disable it when no key is saved. - Use "worker" instead of "pool" in user-facing serverless errors, the serverless guide and the README entry. * feat(serverless): replace Concurrent jobs with a single Images per run setting - Remove the separate Concurrent jobs field; rename Jobs per run to Images per run (1-16) with a tooltip explaining that Runpod's max workers decides how many GPUs run them. - Backend uses a fixed per-worker limit of 16 outstanding jobs, so all images of a run are submitted at once. Unresolved jobs still count against it. The limit is constant, so it never has to change while jobs are unresolved. - Legacy 'concurrency' in saved configs and older browser tabs is accepted and ignored/dropped; it no longer triggers the busy guard. - Revert the enable toggle and show the reason when the backend rejects it. - Trim a pasted api.runpod.ai/v2/<id> URL to the endpoint ID; plainer endpoint-ID error otherwise. - Confirm before a type change deletes a saved Runpod API key. - Update docs and tests.
99 lines
6.0 KiB
Python
99 lines
6.0 KiB
Python
import importlib.util
|
|
import json
|
|
import sys
|
|
import types
|
|
import unittest
|
|
from pathlib import Path
|
|
from unittest.mock import AsyncMock, patch
|
|
|
|
ROOT = Path(__file__).resolve().parents[1]
|
|
PKG = "dist_serverless_status_tests"
|
|
|
|
|
|
class StatusTests(unittest.IsolatedAsyncioTestCase):
|
|
async def asyncSetUp(self):
|
|
def mod(name, **kwargs):
|
|
result = types.ModuleType(name)
|
|
result.__dict__.update(kwargs)
|
|
return result
|
|
self.manager = types.SimpleNamespace(health=AsyncMock(return_value={"ok": True, "queueRemaining": 0, "message": "Ready (scales from zero)"}), public_jobs=lambda _: [], cancel=AsyncMock(return_value=1))
|
|
self.pool = {"id": "pool", "type": "serverless", "endpoint_id": "ep"}
|
|
modules = {PKG: mod(PKG, __path__=[str(ROOT)]), PKG + ".api": mod("api", __path__=[str(ROOT/'api')]),
|
|
PKG + ".utils": mod("utils", __path__=[]),
|
|
PKG + ".utils.config": mod("config", load_config=lambda: {"workers": [self.pool]}),
|
|
PKG + ".api.orchestration": mod("orchestration", __path__=[str(ROOT/'api/orchestration')]),
|
|
PKG + ".api.orchestration.serverless_runtime": mod("runtime", get_manager=lambda: self.manager),
|
|
"server": mod("server", PromptServer=types.SimpleNamespace(instance=types.SimpleNamespace(routes=types.SimpleNamespace(get=lambda _: lambda f: f, post=lambda _: lambda f: f))))}
|
|
with patch.dict(sys.modules, modules):
|
|
spec = importlib.util.spec_from_file_location(PKG+'.api.serverless_routes', ROOT/'api/serverless_routes.py')
|
|
self.routes = importlib.util.module_from_spec(spec)
|
|
spec.loader.exec_module(self.routes)
|
|
|
|
async def test_health_only_never_submits_and_supports_scale_to_zero(self):
|
|
request = types.SimpleNamespace(json=AsyncMock(return_value={"worker_id": "pool"}))
|
|
result = await self.routes.test_serverless_endpoint(request)
|
|
self.assertEqual(result.status, 200)
|
|
self.assertTrue(json.loads(result.text)["ok"])
|
|
self.manager.health.assert_awaited_once_with(self.pool, fresh=True)
|
|
self.manager.cancel.assert_not_awaited()
|
|
|
|
async def test_unknown_pool_cannot_cancel_other_jobs(self):
|
|
request = types.SimpleNamespace(json=AsyncMock(return_value={"worker_id": "other"}))
|
|
result = await self.routes.cancel_serverless_endpoint(request)
|
|
self.assertEqual(result.status, 404)
|
|
self.manager.cancel.assert_not_awaited()
|
|
|
|
async def test_transport_error_does_not_leak_credentials(self):
|
|
self.manager.health.side_effect = RuntimeError("Bearer secret-token https://signed.invalid/?token=value")
|
|
request = types.SimpleNamespace(match_info={"worker_id": "pool"})
|
|
result = await self.routes.serverless_status_endpoint(request)
|
|
self.assertFalse(json.loads(result.text)["ok"])
|
|
self.assertNotIn("secret-token", result.text)
|
|
self.assertNotIn("signed.invalid", result.text)
|
|
|
|
async def test_cancel_uses_only_requested_pool(self):
|
|
request = types.SimpleNamespace(json=AsyncMock(return_value={"worker_id": "pool"}))
|
|
result = await self.routes.cancel_serverless_endpoint(request)
|
|
self.assertEqual(result.status, 200)
|
|
self.manager.cancel.assert_awaited_once_with(worker_id="pool")
|
|
|
|
async def test_reconcile_only_reads_the_requested_pools_jobs(self):
|
|
self.manager.reconcile = AsyncMock(return_value=2)
|
|
request = types.SimpleNamespace(json=AsyncMock(return_value={"worker_id": "pool"}))
|
|
result = await self.routes.reconcile_serverless_endpoint(request)
|
|
self.assertEqual(result.status, 200)
|
|
self.assertEqual(json.loads(result.text)["resolved"], 2)
|
|
self.manager.reconcile.assert_awaited_once_with(worker_id="pool")
|
|
self.manager.cancel.assert_not_awaited()
|
|
|
|
async def test_resolve_requires_explicit_boolean_confirmation_and_record_id(self):
|
|
self.manager.acknowledge_recovery = AsyncMock()
|
|
for payload in ({"worker_id": "pool", "record_id": "record"},
|
|
{"worker_id": "pool", "record_id": "record", "provider_checked": "true"},
|
|
{"worker_id": "pool", "provider_checked": True}):
|
|
with self.subTest(payload=payload):
|
|
result = await self.routes.resolve_serverless_endpoint(types.SimpleNamespace(json=AsyncMock(return_value=payload)))
|
|
self.assertEqual(result.status, 400)
|
|
self.manager.acknowledge_recovery.assert_not_awaited()
|
|
|
|
async def test_resolve_is_scoped_and_does_not_claim_provider_cancellation(self):
|
|
self.manager.acknowledge_recovery = AsyncMock()
|
|
request = types.SimpleNamespace(json=AsyncMock(return_value={"worker_id": "pool", "record_id": "record", "provider_checked": True}))
|
|
result = await self.routes.resolve_serverless_endpoint(request)
|
|
self.assertEqual(result.status, 200)
|
|
self.manager.acknowledge_recovery.assert_awaited_once_with("pool", "record", provider_checked=True)
|
|
self.assertIn("not provider-confirmed", json.loads(result.text)["message"])
|
|
self.manager.cancel.assert_not_awaited()
|
|
|
|
async def test_live_or_mismatched_record_refuses_manual_resolution(self):
|
|
self.manager.acknowledge_recovery = AsyncMock(side_effect=ValueError("Unresolved inactive job not found for this pool"))
|
|
request = types.SimpleNamespace(json=AsyncMock(return_value={"worker_id": "pool", "record_id": "record", "provider_checked": True}))
|
|
result = await self.routes.resolve_serverless_endpoint(request)
|
|
self.assertEqual(result.status, 409)
|
|
|
|
async def test_failed_health_still_exposes_recovery_ids(self):
|
|
self.manager.health.side_effect = RuntimeError("private transport error")
|
|
self.manager.public_jobs = lambda _: [{"record_id": "record", "status": "UNKNOWN"}]
|
|
result = await self.routes.serverless_status_endpoint(types.SimpleNamespace(match_info={"worker_id": "pool"}))
|
|
self.assertEqual(json.loads(result.text)["jobs"][0]["record_id"], "record")
|