diff --git a/validity/scripts/keeper.py b/validity/scripts/keeper.py
index a5b7405..3b7d410 100644
--- a/validity/scripts/keeper.py
+++ b/validity/scripts/keeper.py
@@ -16,8 +16,9 @@ class JobKeeper:
     """
 
     job: Job
-    error_callbacks: list[Callable[[Job], None]] = field(default_factory=list)
+    error_callback: Callable[["JobKeeper", Exception], None] = lambda *_: None  # noqa: E731
     logger: Logger = field(default_factory=lambda: di["Logger"])
+    auto_terminate: bool = True
 
     def __enter__(self):
         self.job.start()
@@ -26,16 +27,11 @@ def __enter__(self):
     def __exit__(self, exc_type, exc, tb):
         if exc_type:
             self.terminate_errored_job(exc)
-        elif self.job.status == JobStatusChoices.STATUS_RUNNING:
+        elif self.job.status == JobStatusChoices.STATUS_RUNNING and self.auto_terminate:
             self.terminate_job()
         self.logger.flush()
 
-    def _exec_callbacks(self):
-        for callback in self.error_callbacks:
-            callback(self)
-
     def terminate_errored_job(self, error: Exception) -> None:
-        self._exec_callbacks()
         if isinstance(error, AbortScript):
             self.logger.messages.extend(error.logs)
             self.logger.failure(str(error))
@@ -43,6 +39,7 @@ def terminate_errored_job(self, error: Exception) -> None:
         else:
             self.logger.log_exception(error)
             status = JobStatusChoices.STATUS_ERRORED
+        self.error_callback(self, error)
         self.terminate_job(status=status, error=repr(error))
 
     def terminate_job(
diff --git a/validity/scripts/runtests/combine.py b/validity/scripts/runtests/combine.py
index 0066caa..278639e 100644
--- a/validity/scripts/runtests/combine.py
+++ b/validity/scripts/runtests/combine.py
@@ -80,12 +80,12 @@ def abort_if_apply_errors(self, job_extractor: JobExtractor) -> None:
             raise AbortScript("ApplyWorkerError", status=JobStatusChoices.STATUS_ERRORED, logs=error_logs)
 
     def get_job_keeper(self, job: Job) -> JobKeeper:
-        def error_callback(keeper):
+        def error_callback(keeper, error):
             keeper.logger.info("Database changes have been reverted")
             self.testresult_queryset.filter(report_id=keeper.job.object_id).raw_delete()
 
         keeper = self.jobkeeper_factory(job)
-        keeper.error_callbacks = [error_callback]
+        keeper.error_callback = error_callback
         return keeper
 
     def __call__(self, params: FullRunTestsParams) -> Any:
diff --git a/validity/scripts/runtests/split.py b/validity/scripts/runtests/split.py
index 14d3c91..7e5bb02 100644
--- a/validity/scripts/runtests/split.py
+++ b/validity/scripts/runtests/split.py
@@ -1,4 +1,5 @@
 from dataclasses import dataclass, field
+from functools import partial
 from itertools import chain, cycle, groupby, repeat
 from typing import Any, Callable, Collection, Iterable, Protocol
 
@@ -25,7 +26,7 @@ def __call__(
 @di.dependency(scope=Singleton)
 @dataclass(repr=False)
 class SplitWorker:
-    jobkeeper_factory: Callable[[Job], JobKeeper] = JobKeeper
+    jobkeeper_factory: Callable[[Job], JobKeeper] = partial(JobKeeper, auto_terminate=False)
     datasource_sync_fn: Callable[[Iterable[VDataSource], Q], None] = datasource_sync
     backup_fn: BackupFn = bulk_backup
     device_batch_size: int = 2000
diff --git a/validity/tests/test_scripts/runtests/test_apply.py b/validity/tests/test_scripts/runtests/test_apply.py
index b6b4612..a83d405 100644
--- a/validity/tests/test_scripts/runtests/test_apply.py
+++ b/validity/tests/test_scripts/runtests/test_apply.py
@@ -11,6 +11,7 @@
 from validity.scripts.data_models import TestResultRatio as ResultRatio
 from validity.scripts.runtests.apply import ApplyWorker, DeviceTestIterator
 from validity.scripts.runtests.apply import TestExecutor as TExecutor
+from validity.utils.logger import Logger
 
 
 NS_1 = """
@@ -48,7 +49,7 @@ def func3(): pass
 )
 @pytest.mark.django_db
 def test_nameset_functions(nameset_texts, extracted_fn_names, warning_calls):
-    script = TExecutor(1, 2, 10)
+    script = TExecutor(Logger(), 2, 10)
     namesets = [NameSetDBFactory(definitions=d) for d in nameset_texts]
     functions = script.nameset_functions(namesets)
     assert extracted_fn_names == functions.keys()
@@ -73,7 +74,7 @@ def test_nameset_functions(nameset_texts, extracted_fn_names, warning_calls):
 )
 @pytest.mark.django_db
 def test_builtins_are_available_in_nameset(definitions):
-    script = TExecutor(10, 20, 30, extra_globals=DEFAULT_NAMESET)
+    script = TExecutor(Logger(), 20, 30, extra_globals=DEFAULT_NAMESET)
     namesets = [NameSetDBFactory(definitions=definitions)]
     functions = script.nameset_functions(namesets)
     functions["func"]()
@@ -89,7 +90,7 @@ def test_run_tests_for_device():
     tests[0].run.return_value = True, []
     tests[1].run.return_value = False, [("some", "explanation")]
     tests[2].run.side_effect = EvalError("some test error")
-    executor = TExecutor(10, explanation_verbosity=2, report_id=30)
+    executor = TExecutor(Logger(), explanation_verbosity=2, report_id=30)
     results = [
         {
             "passed": r.passed,
@@ -132,6 +133,7 @@ def apply_worker():
     job_extractor_factory = Mock()
     job_extractor_factory.return_value.parent.job.result.slices = [None, {1: [1, 2, 3]}]
     return ApplyWorker(
+        logger=Logger(),
         testresult_queryset=Mock(),
         test_executor_factory=Mock(return_value=executor),
         result_batch_size=100,
@@ -160,8 +162,7 @@ def test_applyworker_success(full_runtests_params, apply_worker):
 
 
 @dataclass
-class MockLogger:
-    script_id: str
+class MockLogger(Logger):
     messages: list = field(default_factory=list, init=False)
 
     def log_exception(self, m):
@@ -171,6 +172,6 @@ def log_exception(self, m):
 @pytest.mark.django_db
 def test_applyworker_exception(full_runtests_params, apply_worker):
     apply_worker.test_executor_factory = Mock(side_effect=ValueError("some error"))
-    apply_worker.logger_factory = MockLogger
+    apply_worker.logger = MockLogger()
     result = apply_worker(params=full_runtests_params, worker_id=1)
     assert result == ExecutionResult(test_stat=ResultRatio(passed=0, total=0), log=["some error"], errored=True)
diff --git a/validity/tests/test_scripts/runtests/test_combine.py b/validity/tests/test_scripts/runtests/test_combine.py
index 09d20c8..77d3ff3 100644
--- a/validity/tests/test_scripts/runtests/test_combine.py
+++ b/validity/tests/test_scripts/runtests/test_combine.py
@@ -9,6 +9,7 @@
 from validity.scripts.data_models import TestResultRatio as ResultRatio
 from validity.scripts.exceptions import AbortScript
 from validity.scripts.runtests.combine import CombineWorker
+from validity.utils.logger import Logger
 
 
 @pytest.fixture
@@ -38,9 +39,8 @@ def job_extractor(messages):
 # but according to netbox4.0 strange behaviour reverse() finally causes it
 @pytest.mark.django_db
 def test_compose_logs(worker, messages, job_extractor):
-    logger = worker.log_factory()
     time = messages[0].time
-    logs = worker.compose_logs(logger, job_extractor, report_id=10)
+    logs = worker.compose_logs(Logger(), job_extractor, report_id=10)
     assert len(logs) == 6
     assert logs[:5] == messages
     last_msg = replace(logs[-1], time=time)
diff --git a/validity/tests/test_scripts/runtests/test_split.py b/validity/tests/test_scripts/runtests/test_split.py
index 359d071..0a56b5c 100644
--- a/validity/tests/test_scripts/runtests/test_split.py
+++ b/validity/tests/test_scripts/runtests/test_split.py
@@ -10,6 +10,7 @@
 
 from validity.scripts.data_models import Message, SplitResult
 from validity.scripts.runtests.split import SplitWorker
+from validity.utils.logger import Logger
 
 
 @pytest.fixture
@@ -50,9 +51,7 @@ def split_worker():
 def test_distribute_work(split_worker, selectors, worker_num, runtests_params, expected_result, devices):
     runtests_params.workers_num = worker_num
     runtests_params.selectors = [s.pk for s in selectors]
-    result = split_worker.distribute_work(
-        runtests_params, split_worker.log_factory(), runtests_params.get_device_filter()
-    )
+    result = split_worker.distribute_work(runtests_params, Logger(), runtests_params.get_device_filter())
     assert result == expected_result
 
 
@@ -103,13 +102,11 @@ def test_call(selectors, devices, runtests_params, monkeypatch):
                 status="info",
                 message="Running the tests for *2 devices*",
                 time=datetime.datetime(2000, 1, 1, 0, 0),
-                script_id=None,
             ),
             Message(
                 status="info",
                 message="Distributing the work among 2 workers. Each worker handles 1 device(s) in average",
                 time=datetime.datetime(2000, 1, 1, 0, 0),
-                script_id=None,
             ),
         ],
         slices=[{1: [1]}, {2: [2]}],
diff --git a/validity/tests/test_scripts/test_launcher.py b/validity/tests/test_scripts/test_launcher.py
index dddf39d..a52c1e4 100644
--- a/validity/tests/test_scripts/test_launcher.py
+++ b/validity/tests/test_scripts/test_launcher.py
@@ -1,5 +1,4 @@
 import uuid
-from dataclasses import asdict
 from unittest.mock import Mock
 
 import pytest
@@ -13,8 +12,8 @@
 
 
 class ConcreteScriptParams(ScriptParams):
-    def with_job_info(self, job: Job):
-        return FullParams(**asdict(self) | {"job": job})
+    def _full_cls(self):
+        return FullParams
 
 
 class FullParams:
@@ -28,7 +27,13 @@ def __eq__(self, other):
 @pytest.fixture
 def launcher(db):
     report = ComplianceReport.objects.create()
-    return Launcher(job_name="test_launcher", job_object_factory=lambda: report, rq_queue=Mock(), tasks=[])
+    return Launcher(
+        job_name="test_launcher",
+        job_object_factory=lambda _: report,
+        rq_queue=Mock(),
+        tasks=[],
+        worker_count_fn=lambda _: 3,
+    )
 
 
 @pytest.fixture
@@ -45,7 +50,7 @@ def task_func(): ...
     params.schedule_at = schedule_at
     launcher.tasks = [Task(task_func, job_timeout=60)]
     job = launcher(params)
-    assert isinstance(job, Job) and job.object == launcher.job_object_factory()
+    assert isinstance(job, Job) and job.object == launcher.job_object_factory(None)
     enqueue_fn = getattr(launcher.rq_queue, "enqueue_at" if schedule_at else "enqueue")
     enqueue_fn.assert_called_once()
     enqueue_kwargs = enqueue_fn.call_args.kwargs
diff --git a/validity/tests/test_scripts/test_logger.py b/validity/tests/test_utils/test_logger.py
similarity index 88%
rename from validity/tests/test_scripts/test_logger.py
rename to validity/tests/test_utils/test_logger.py
index 4c6bb5e..41ebd00 100644
--- a/validity/tests/test_scripts/test_logger.py
+++ b/validity/tests/test_utils/test_logger.py
@@ -1,6 +1,6 @@
 import pytest
 
-from validity.scripts.logger import Logger
+from validity.utils.logger import Logger
 
 
 @pytest.fixture
@@ -29,7 +29,7 @@ def test_logger(error_with_traceback):
             "status": "failure",
             "message": (
                 "Unhandled error occured: `<class 'ValueError'>: error`\n```\n  "
-                'File "/plugin/validity/validity/tests/test_scripts/test_logger.py", '
+                f'File "{__file__}", '
                 """line 10, in error_with_traceback\n    raise ValueError("error")\n\n```"""
             ),
         },
diff --git a/validity/tests/test_views.py b/validity/tests/test_views.py
index aa981a0..faef344 100644
--- a/validity/tests/test_views.py
+++ b/validity/tests/test_views.py
@@ -218,18 +218,16 @@ def test_get(self, admin_client):
         assert resp.status_code == HTTPStatus.OK
 
     @pytest.mark.parametrize(
-        "form_data, status_code, worker_count",
+        "form_data, status_code, has_workers",
         [
-            ({}, HTTPStatus.FOUND, 1),
-            ({}, HTTPStatus.OK, 0),
-            ({"devices": [1, 2]}, HTTPStatus.OK, 1),  # devices do not exist
+            ({}, HTTPStatus.FOUND, True),
+            ({}, HTTPStatus.OK, False),
+            ({"devices": [1, 2]}, HTTPStatus.OK, True),  # devices do not exist
         ],
     )
-    def test_post(self, admin_client, di, form_data, status_code, worker_count):
-        launcher = Mock(**{"rq_queue.name": "queue_1", "return_value.pk": 1})
-        with di.override(
-            {dependencies.runtests_launcher: lambda: launcher, dependencies.runtests_worker_count: lambda: worker_count}
-        ):
+    def test_post(self, admin_client, di, form_data, status_code, has_workers):
+        launcher = Mock(**{"has_workers": has_workers, "return_value.pk": 1})
+        with di.override({dependencies.runtests_launcher: lambda: launcher}):
             result = admin_client.post(self.url, form_data)
             assert result.status_code == status_code
             if status_code == HTTPStatus.FOUND:  # if form is valid