Python test_utils.do_job_ingestion函数代码示例

def test_cycle_one_job(jm, refdata, sample_data, initial_data, sample_resultset, mock_log_parser):
    """
    Test cycling one job in a group of jobs to confirm there are no
    unexpected deletions
    """

    job_data = sample_data.job_data[:20]
    test_utils.do_job_ingestion(jm, refdata, job_data, sample_resultset, False)

    time_now = time.time()
    cycle_date_ts = int(time_now - 7 * 24 * 3600)

    jm.execute(proc="jobs_test.updates.set_result_sets_push_timestamp", placeholders=[time_now])

    jm.execute(proc="jobs_test.updates.set_one_result_set_push_timestamp", placeholders=[cycle_date_ts])

    jobs_to_be_deleted = jm.execute(proc="jobs_test.selects.get_result_set_jobs", placeholders=[1])

    jobs_before = jm.execute(proc="jobs_test.selects.jobs")

    call_command("cycle_data", sleep_time=0, cycle_interval=1, debug=True)

    jobs_after = jm.execute(proc="jobs_test.selects.jobs")

    # Confirm that the target result set has no jobs in the
    # jobs table
    jobs_to_be_deleted_after = jm.execute(proc="jobs_test.selects.get_result_set_jobs", placeholders=[1])

    assert len(jobs_to_be_deleted_after) == 0

    assert len(jobs_after) == len(jobs_before) - len(jobs_to_be_deleted)

def test_ingest_all_sample_jobs(jm, sample_data,
                                sample_resultset, test_repository, mock_log_parser):
    """
    Process each job structure in the job_data.txt file and verify.
    """
    job_data = sample_data.job_data
    test_utils.do_job_ingestion(jm, job_data, sample_resultset)

def test_cycle_all_data_in_chunks(test_repository, failure_classifications, sample_data,
                                  sample_resultset, mock_log_parser):
    """
    Test cycling the sample data in chunks.
    """
    job_data = sample_data.job_data[:20]
    test_utils.do_job_ingestion(test_repository, job_data, sample_resultset, False)

    # build a date that will cause the data to be cycled
    cycle_date_ts = datetime.datetime.now() - datetime.timedelta(weeks=1)
    for job in Job.objects.all():
        job.submit_time = cycle_date_ts
        job.save()

    create_failure_lines(Job.objects.get(id=1),
                         [(test_line, {})] * 7)

    assert TestFailureLine.search().count() > 0

    call_command('cycle_data', sleep_time=0, days=1, chunk_size=3)
    refresh_all()

    # There should be no jobs after cycling
    assert Job.objects.count() == 0
    assert FailureLine.objects.count() == 0
    assert JobDetail.objects.count() == 0
    assert TestFailureLine.search().count() == 0

def test_cycle_job_model_reference_data(test_repository, failure_classifications,
                                        sample_data, sample_resultset,
                                        mock_log_parser):
    job_data = sample_data.job_data[:20]
    test_utils.do_job_ingestion(test_repository, job_data, sample_resultset, False)

    # get a list of ids of original reference data
    original_job_type_ids = JobType.objects.values_list('id', flat=True)
    original_job_group_ids = JobGroup.objects.values_list('id', flat=True)
    original_machine_ids = Machine.objects.values_list('id', flat=True)

    # create a bunch of job model data that should be cycled, since they don't
    # reference any current jobs
    jg = JobGroup.objects.create(symbol='moo', name='moo')
    jt = JobType.objects.create(job_group=jg, symbol='mu', name='mu')
    m = Machine.objects.create(name='machine_with_no_job')
    (jg_id, jt_id, m_id) = (jg.id, jt.id, m.id)
    call_command('cycle_data', sleep_time=0, days=1, chunk_size=3)

    # assert that reference data that should have been cycled, was cycled
    assert JobGroup.objects.filter(id=jg_id).count() == 0
    assert JobType.objects.filter(id=jt_id).count() == 0
    assert Machine.objects.filter(id=m_id).count() == 0

    # assert that we still have everything that shouldn't have been cycled
    assert JobType.objects.filter(id__in=original_job_type_ids).count() == len(original_job_type_ids)
    assert JobGroup.objects.filter(id__in=original_job_group_ids).count() == len(original_job_group_ids)
    assert Machine.objects.filter(id__in=original_machine_ids).count() == len(original_machine_ids)

def test_cycle_all_data(jm, refdata, sample_data, initial_data,
                        sample_resultset, mock_log_parser):
    """
    Test cycling the sample data
    """
    job_data = sample_data.job_data[:20]
    test_utils.do_job_ingestion(jm, refdata, job_data, sample_resultset, False)

    time_now = time.time()
    cycle_date_ts = time_now - 7 * 24 * 3600

    jm.execute(
        proc="jobs_test.updates.set_result_sets_push_timestamp",
        placeholders=[cycle_date_ts]
    )

    jobs_to_be_deleted = jm.execute(
        proc="jobs_test.selects.get_jobs_for_cycling",
        placeholders=[time_now - 24 * 3600]
    )

    jobs_before = jm.execute(proc="jobs_test.selects.jobs")

    call_command('cycle_data', sleep_time=0, cycle_interval=1)

    jobs_after = jm.execute(proc="jobs_test.selects.jobs")

    assert len(jobs_after) == len(jobs_before) - len(jobs_to_be_deleted)

    # There should be no jobs after cycling
    assert len(jobs_after) == 0

def test_cycle_all_data(jm, sample_data,
                        sample_resultset, test_repository, mock_log_parser,
                        failure_lines):
    """
    Test cycling the sample data
    """
    job_data = sample_data.job_data[:20]
    test_utils.do_job_ingestion(jm, job_data, sample_resultset, False)

    time_now = time.time()
    cycle_date_ts = time_now - 7 * 24 * 3600

    jm.execute(
        proc="jobs_test.updates.set_jobs_last_modified",
        placeholders=[cycle_date_ts]
    )

    jobs_to_be_deleted = jm.execute(
        proc="jobs_test.selects.get_jobs_for_cycling",
        placeholders=[time_now - 24 * 3600]
    )

    jobs_before = jm.execute(proc="jobs_test.selects.jobs")

    call_command('cycle_data', sleep_time=0, days=1)

    jobs_after = jm.execute(proc="jobs_test.selects.jobs")

    assert len(jobs_after) == len(jobs_before) - len(jobs_to_be_deleted)

    # There should be no jobs or failure lines after cycling
    assert len(jobs_after) == 0
    assert FailureLine.objects.count() == 0
    assert Job.objects.count() == 0

def test_cycle_all_data(test_repository, failure_classifications, sample_data,
                        sample_resultset, mock_log_parser, failure_lines):
    """
    Test cycling the sample data
    """
    job_data = sample_data.job_data[:20]
    test_utils.do_job_ingestion(test_repository, job_data, sample_resultset, False)

    # set the submit time to be a week before today
    cycle_date_ts = datetime.datetime.now() - datetime.timedelta(weeks=1)
    for job in Job.objects.all():
        job.submit_time = cycle_date_ts
        job.save()

    call_command('cycle_data', sleep_time=0, days=1)

    refresh_all()

    # There should be no jobs or failure lines after cycling
    assert Job.objects.count() == 0
    assert FailureLine.objects.count() == 0
    assert JobDetail.objects.count() == 0
    assert JobLog.objects.count() == 0

    # There should be nothing in elastic search after cycling
    assert TestFailureLine.search().count() == 0

def test_ingest_all_sample_jobs(test_repository, failure_classifications,
                                sample_data, sample_push, mock_log_parser):
    """
    Process each job structure in the job_data.txt file and verify.
    """
    job_data = sample_data.job_data
    test_utils.do_job_ingestion(test_repository, job_data, sample_push)

def test_cycle_all_data_in_chunks(jm, refdata, sample_data, initial_data, sample_resultset, mock_log_parser):
    """
    Test cycling the sample data in chunks.
    """
    job_data = sample_data.job_data[:20]
    test_utils.do_job_ingestion(jm, refdata, job_data, sample_resultset, False)

    # build a date that will cause the data to be cycled
    time_now = time.time()
    cycle_date_ts = int(time_now - 7 * 24 * 3600)

    jm.execute(proc="jobs_test.updates.set_result_sets_push_timestamp", placeholders=[cycle_date_ts])

    jobs_to_be_deleted = jm.execute(proc="jobs_test.selects.get_jobs_for_cycling", placeholders=[time_now - 24 * 3600])

    jobs_before = jm.execute(proc="jobs_test.selects.jobs")

    call_command("cycle_data", sleep_time=0, cycle_interval=1, chunk_size=3)

    jobs_after = jm.execute(proc="jobs_test.selects.jobs")

    assert len(jobs_after) == len(jobs_before) - len(jobs_to_be_deleted)

    # There should be no jobs after cycling
    assert len(jobs_after) == 0

def test_ingest_single_sample_job(jm, refdata, sample_data, initial_data, mock_log_parser, sample_resultset):
    """Process a single job structure in the job_data.txt file"""
    job_data = sample_data.job_data[:1]
    test_utils.do_job_ingestion(jm, refdata, job_data, sample_resultset)

    jm.disconnect()
    refdata.disconnect()

def test_ingesting_skip_existing(test_repository, failure_classifications, sample_data,
                                 sample_push, mock_log_parser):
    """Remove single existing job prior to loading"""
    job_data = sample_data.job_data[:1]
    test_utils.do_job_ingestion(test_repository, job_data, sample_push)

    store_job_data(test_repository, sample_data.job_data[:2])

    assert Job.objects.count() == 2

def test_cycle_one_job(jm, sample_data,
                       sample_resultset, test_repository, mock_log_parser,
                       failure_lines):
    """
    Test cycling one job in a group of jobs to confirm there are no
    unexpected deletions
    """

    job_data = sample_data.job_data[:20]
    test_utils.do_job_ingestion(jm, job_data, sample_resultset, False)

    job_not_deleted = jm.get_job(2)[0]

    failure_lines_remaining = create_failure_lines(test_repository,
                                                   job_not_deleted["job_guid"],
                                                   [(test_line, {}),
                                                    (test_line, {"subtest": "subtest2"})])

    time_now = time.time()
    cycle_date_ts = int(time_now - 7 * 24 * 3600)

    jm.execute(
        proc="jobs_test.updates.set_jobs_last_modified",
        placeholders=[time_now]
    )

    jm.execute(
        proc="jobs_test.updates.set_one_job_last_modified_timestamp",
        placeholders=[cycle_date_ts]
    )

    jobs_to_be_deleted = jm.execute(
        proc="jobs_test.selects.get_one_job_for_cycling",
        placeholders=[1]
    )

    jobs_before = jm.execute(proc="jobs_test.selects.jobs")

    call_command('cycle_data', sleep_time=0, days=1, debug=True)

    jobs_after = jm.execute(proc="jobs_test.selects.jobs")

    # Confirm that the target result set has no jobs in the
    # jobs table
    jobs_to_be_deleted_after = jm.execute(
        proc="jobs_test.selects.get_one_job_for_cycling",
        placeholders=[1]
    )

    assert len(jobs_to_be_deleted_after) == 0

    assert len(jobs_after) == len(jobs_before) - len(jobs_to_be_deleted)
    assert len(jobs_after) == Job.objects.count()

    assert (set(item.id for item in FailureLine.objects.all()) ==
            set(item.id for item in failure_lines_remaining))

def test_get_job_data(jm, test_project, refdata, sample_data, initial_data, mock_log_parser, sample_resultset):

    target_len = 10
    job_data = sample_data.job_data[:target_len]
    test_utils.do_job_ingestion(jm, refdata, job_data, sample_resultset)

    with ArtifactsModel(test_project) as artifacts_model:
        job_data = artifacts_model.get_job_signatures_from_ids(range(1, 11))

    assert len(job_data) is target_len

def test_get_job_data(jm, refdata, sample_data, initial_data,
                                  mock_log_parser, sample_resultset):

    target_len = 10
    job_data = sample_data.job_data[:target_len]
    test_utils.do_job_ingestion(jm, refdata, job_data, sample_resultset)

    job_data = jm.get_job_signatures_from_ids(range(1,11))

    assert len(job_data) is target_len

def test_ingesting_skip_existing(jm, sample_data, initial_data, refdata, mock_log_parser, sample_resultset):
    """Remove single existing job prior to loading"""

    job_data = sample_data.job_data[:1]
    test_utils.do_job_ingestion(jm, refdata, job_data, sample_resultset)

    jm.store_job_data(sample_data.job_data[:2])

    jl = jm.get_job_list(0, 10)
    assert len(jl) == 2