class ChecksumDaemonTest(UploadTestCaseUsingMockAWS):
def _make_dbfile(self, upload_area, test_file, checksums=None):
return DbFile(s3_key=f"{upload_area.uuid}/{test_file.name}", s3_etag=test_file.e_tag,
upload_area_id=upload_area.db_id, name=test_file.name, size=test_file.size,
checksums=checksums)
def setUp(self):
super().setUp()
# Environment
self.environment = {
'INGEST_AMQP_SERVER': 'foo',
'CSUM_DOCKER_IMAGE': 'bogoimage'
}
self.environmentor = EnvironmentSetup(self.environment)
self.environmentor.enter()
# Upload area
self.area_uuid = str(uuid.uuid4())
self.upload_area = UploadArea(self.area_uuid)
self.upload_area.update_or_create()
# daemon
context = Mock()
self.daemon = ChecksumDaemon(context)
# File
self.small_file = FixtureFile.factory('foo')
self.file_key = f"{self.area_uuid}/{self.small_file.name}"
self.object = self.upload_bucket.Object(self.file_key)
self.object.put(Key=self.file_key, Body=self.small_file.contents, ContentType=self.small_file.content_type)
# Event
self.events = {'Records': [
{'eventVersion': '2.0', 'eventSource': 'aws:s3', 'awsRegion': 'us-east-1',
'eventTime': '2017-09-15T00:05:10.378Z', 'eventName': 'ObjectCreated:Put',
'userIdentity': {'principalId': 'AWS:AROAI4WRRXW2K3Y2IFL6Q:upload-api-dev'},
'requestParameters': {'sourceIPAddress': '52.91.56.220'},
'responseElements': {'x-amz-request-id': 'FEBC85CADD1E3A66',
'x-amz-id-2': 'xxx'},
's3': {'s3SchemaVersion': '1.0',
'configurationId': 'NGZjNmM0M2ItZTk0Yi00YTExLWE2NDMtMzYzY2UwN2EyM2Nj',
'bucket': {'name': self.upload_config.bucket_name,
'ownerIdentity': {'principalId': 'A29PZ5XRQWJUUM'},
'arn': f'arn:aws:s3:::{self.upload_config.bucket_name}'},
'object': {'key': self.file_key,
'size': self.small_file.size,
'eTag': self.small_file.e_tag,
'sequencer': '0059BB193641C4EAB0'}}}]}
self.db_session_maker = DBSessionMaker()
self.db = self.db_session_maker.session()
class TestUploadedFile(UploadTestCaseUsingMockAWS):
def setUp(self):
super().setUp()
self.db_session_maker = DBSessionMaker()
self.db = self.db_session_maker.session()
self.upload_area_id = str(uuid.uuid4())
self.upload_area = UploadArea(self.upload_area_id)
self.upload_area.update_or_create()
def create_file_record(self, s3object, checksums=None):
record = DbFile(s3_key=s3object.key,
s3_etag=s3object.e_tag.strip('\"'),
name=os.path.basename(s3object.key),
upload_area_id=self.upload_area.db_id,
size=s3object.content_length,
checksums=checksums)
self.db.add(record)
self.db.commit()
return record
def tearDown(self):
super().tearDown()
def test_create__creates_a_new_s3_object_and_db_record(self):
filename = f"file-{random.randint(0, 999999999)}"
content_type = "application/octet-stream; dcp-type=data"
file_content = "file1_content"
uf = UploadedFile.create(upload_area=self.upload_area,
name=filename,
content_type=content_type,
data=file_content)
self.assertIsInstance(uf, UploadedFile)
# S3 Object
s3_key = f"{self.upload_area_id}/{filename}"
s3object = self.upload_bucket.Object(s3_key)
self.assertEqual(content_type, s3object.content_type)
self.assertEqual(file_content.encode('utf8'), s3object.get()['Body'].read())
# DB Record
record = self.db.query(DbFile).filter(DbFile.s3_key == s3_key,
DbFile.s3_etag == s3object.e_tag.strip('\"')).one()
self.assertEqual(s3_key, record.s3_key)
self.assertEqual(filename, record.name)
self.assertEqual(s3object.e_tag.strip('\"'), record.s3_etag)
self.assertEqual(len(file_content), record.size)
self.assertEqual(self.upload_area.db_id, record.upload_area_id)
def test_init__given_existing_entities__initializes_properties_correctly(self):
filename = f"file-{random.randint(0, 999999999)}"
s3object = self.create_s3_object(f"{self.upload_area_id}/{filename}")
file_record = self.create_file_record(s3object)
uf = UploadedFile(self.upload_area, s3object=s3object)
# Links to objects
self.assertEqual(s3object, uf.s3object)
self.assertEqual(self.upload_area, uf.upload_area)
# Persisted properties
self.assertEqual(file_record.id, uf.db_id)
self.assertEqual(s3object.key, uf.s3_key)
self.assertEqual(s3object.e_tag.strip('\"'), uf.s3_etag)
self.assertEqual(self.upload_area.db_id, uf._properties['upload_area_id'])
self.assertEqual(file_record.name, uf.name)
self.assertEqual(s3object.content_length, uf.size)
def test_init__when_no_db_record_exists__creates_a_db_record(self):
filename = f"file-{random.randint(0, 999999999)}"
s3object = self.create_s3_object(f"{self.upload_area_id}/{filename}")
with self.assertRaises(NoResultFound):
self.db.query(DbFile).filter(DbFile.s3_key == s3object.key,
DbFile.s3_etag == s3object.e_tag.strip('\"')).one()
uf = UploadedFile(upload_area=self.upload_area, s3object=s3object)
record = self.db.query(DbFile).filter(DbFile.s3_key == s3object.key,
DbFile.s3_etag == s3object.e_tag.strip('\"')).one()
self.assertEqual(record.id, uf.db_id)
self.assertEqual(s3object.key, record.s3_key)
self.assertEqual(filename, record.name)
self.assertEqual(s3object.e_tag.strip('\"'), record.s3_etag)
self.assertEqual(s3object.content_length, record.size)
self.assertEqual(self.upload_area.db_id, record.upload_area_id)
def test_init__doesnt_create_db_record_if_one_already_exists(self):
filename = f"file-{random.randint(0, 999999999)}"
s3_key = f"{self.upload_area_id}/{filename}"
s3object = self.create_s3_object(s3_key)
self.create_file_record(s3object)
record_count_before = self.db.query(DbFile).filter(DbFile.s3_key == s3_key).count()
UploadedFile(upload_area=self.upload_area, s3object=s3object)
record_count_after = self.db.query(DbFile).filter(DbFile.s3_key == s3_key).count()
self.assertEqual(record_count_before, record_count_after)
def test_from_s3_key__initializes_correctly(self):
filename = f"file-{random.randint(0, 999999999)}"
s3object = self.create_s3_object(f"{self.upload_area_id}/{filename}")
file_record = self.create_file_record(s3object)
uf = UploadedFile.from_s3_key(self.upload_area, s3_key=s3object.key)
self.assertEqual(self.upload_area, uf.upload_area)
self.assertEqual(s3object, uf.s3object)
self.assertEqual(file_record.id, uf.db_id)
def test_from_db_id__initializes_correctly_and_figures_out_which_upload_area_to_use(self):
filename = f"file-{random.randint(0, 999999999)}"
s3object = self.create_s3_object(f"{self.upload_area_id}/{filename}")
file_record = self.create_file_record(s3object)
uf = UploadedFile.from_db_id(file_record.id)
self.assertEqual(self.upload_area.uuid, uf.upload_area.uuid)
self.assertEqual(self.upload_area.db_id, uf.upload_area.db_id)
self.assertEqual(s3object, uf.s3object)
self.assertEqual(file_record.id, uf.db_id)
def test_refresh__picks_up_changed_content_type(self):
filename = f"file-{random.randint(0, 999999999)}"
old_content_type = "application/octet-stream" # missing dcp-type
new_content_type = "application/octet-stream; dcp-type=data"
s3object = self.create_s3_object(object_key=f"{self.upload_area.uuid}/{filename}",
content_type=old_content_type)
# create UploadedFile
uf = UploadedFile.from_s3_key(upload_area=self.upload_area, s3_key=s3object.key)
# Change media type on S3 object
s3object.copy_from(CopySource={'Bucket': self.upload_config.bucket_name, 'Key': s3object.key},
MetadataDirective="REPLACE",
ContentType=new_content_type)
self.assertEqual(old_content_type, uf.content_type)
uf.refresh()
self.assertEqual(new_content_type, uf.content_type)
def test_checksums_setter_saves_db_record(self):
filename = f"file-{random.randint(0, 999999999)}"
s3object = self.create_s3_object(f"{self.upload_area_id}/{filename}")
file_record = self.create_file_record(s3object)
uf = UploadedFile.from_db_id(file_record.id)
uf.checksums = {'foo': 'bar'}
self.db.refresh(file_record)
self.assertEqual({'foo': 'bar'}, file_record.checksums)
def test_info(self):
test_file = FixtureFile.factory("foo")
s3object = self.create_s3_object(f"{self.upload_area_id}/foo", content=test_file.contents)
file_record = self.create_file_record(s3object, checksums=test_file.checksums)
uf = UploadedFile(self.upload_area, s3object=s3object)
self.assertEqual({
'upload_area_id': self.upload_area.uuid,
'name': file_record.name,
'size': s3object.content_length,
'content_type': s3object.content_type,
'url': f"s3://{s3object.bucket_name}/{s3object.key}",
'checksums': test_file.checksums,
'last_modified': s3object.last_modified.isoformat()
}, uf.info())
class UploadAreaTest(UploadTestCaseUsingMockAWS):
def setUp(self):
super().setUp()
self.db_session_maker = DBSessionMaker()
self.db = self.db_session_maker.session()
class TestUploadService(unittest.TestCase):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.batch = boto3.client('batch')
self.uri = None
self.db_session_maker = DBSessionMaker()
def setUp(self):
_start_time = time.time()
self.api_url = f"https://{os.environ['API_HOST']}/v1"
self.upload_config = UploadConfig()
self.auth_headers = {'Api-Key': self.upload_config.api_key}
self.deployment_stage = os.environ['DEPLOYMENT_STAGE']
self.upload_area_uuid = "deadbeef-dead-dead-dead-%012d" % random.randint(
0, 999999999999)
self.verbose = True
_end_time = time.time()
print(f"Total startup time: {_end_time - _start_time} seconds.")
def test__upload_small_file__successful(self):
# Test variables
_start_time = time.time()
_small_file = FixtureFile.factory('small_file')
# Run test
print(
f"\n\nUsing environment {self.deployment_stage} at URL {self.api_url}.\n"
)
self._execute_create_upload_area()
self._execute_upload_file_using_cli(_small_file.path)
self._verify_file_was_checksummed_inline(_small_file)
_validation_id = self._execute_validate_file(_small_file)
self._verify_file_validation_status(
_validation_id
) # default parameters checks for success in validation
self._execute_forget_upload_area()
self._execute_delete_upload_area()
_end_time = time.time()
print(
f"Total test_upload__small_file__successful time: {_end_time - _start_time} seconds."
)
def test__upload_large_file__successful(self):
# Test variables
_start_time = time.time()
_large_file = FixtureFile.factory('10241MB_file')
# Run test
print(
f"\n\nUsing environment {self.deployment_stage} at URL {self.api_url}.\n"
)
self._execute_create_upload_area()
self._execute_upload_file_using_cli(_large_file.url)
self._verify_file_is_checksummed_via_batch(_large_file)
self._execute_forget_upload_area()
self._execute_delete_upload_area()
_end_time = time.time()
print(
f"Total test__upload_large_file__successful time: {_end_time - _start_time} seconds."
)
def test__upload_invalid_file__validation_result_shows_invalid_state(self):
# Test variables
_start_time = time.time()
_invalid_file = FixtureFile.factory('small_invalid_file')
# Run test
print(
f"\n\nUsing environment {self.deployment_stage} at URL {self.api_url}.\n"
)
self._execute_create_upload_area()
self._execute_upload_file_using_cli(_invalid_file.path)
self._verify_file_was_checksummed_inline(_invalid_file)
_validation_id = self._execute_validate_file(_invalid_file)
# Verify that the validation result of the file is invalid. This is designated by an exit code of 1 and the
# presence of an error message saying that file is invalid.
self._verify_file_validation_status(_validation_id, 1, "invalid")
self._execute_forget_upload_area()
self._execute_delete_upload_area()
_end_time = time.time()
print(
f"Total test__upload_invalid_file__validation_result_shows_invalid_state time: {_end_time - _start_time} "
f"seconds.")
def _execute_create_upload_area(self):
response = self._make_request(
description="CREATE UPLOAD AREA",
verb='POST',
url=f"{self.api_url}/area/{self.upload_area_uuid}",
headers=self.auth_headers,
expected_status=201)
data = json.loads(response)
self.uri = data['uri']
self.assertEqual('UNLOCKED', self._get_upload_area_record_status())
def _execute_upload_file_using_cli(self, file_location):
self._run_cli_command("SELECT UPLOAD AREA",
['hca', 'upload', 'select', self.uri])
self._run_cli_command("UPLOAD FILE USING CLI",
['hca', 'upload', 'files', file_location])
def _execute_validate_file(self, test_file):
response = self._make_request(
description="VALIDATE",
verb='PUT',
url=
f"{self.api_url}/area/{self.upload_area_uuid}/{test_file.name}/validate",
expected_status=200,
headers=self.auth_headers,
json={
"validator_image": "humancellatlas/upload-validator-example:14"
})
validation_id = json.loads(response)['validation_id']
WaitFor(self._get_validation_record_status, validation_id) \
.to_return_value('SCHEDULED', timeout_seconds=MINUTE_SEC)
validation_job_id = self._get_validation_record_job_id(validation_id)
WaitFor(self._get_batch_job_status, validation_job_id) \
.to_return_value('SUCCEEDED', timeout_seconds=20 * MINUTE_SEC)
WaitFor(self._get_validation_record_status, validation_id) \
.to_return_value('VALIDATED', timeout_seconds=MINUTE_SEC)
return validation_id
def _execute_forget_upload_area(self):
self._run_cli_command(
"FORGET UPLOAD AREA",
['hca', 'upload', 'forget', self.upload_area_uuid])
def _execute_delete_upload_area(self):
self._make_request(description="DELETE UPLOAD AREA",
verb='DELETE',
url=f"{self.api_url}/area/{self.upload_area_uuid}",
headers=self.auth_headers,
expected_status=202)
WaitFor(self._get_upload_area_record_status) \
.to_return_value('DELETED', timeout_seconds=MINUTE_SEC)
def _verify_file_was_checksummed_inline(self, test_file):
""" For files that are smaller than 10G, we expect that the file will be check-summed inline. This means that
there is no need to schedule a job in batch and no job id is given to the checksum record."""
print("VERIFYING FILE WAS CHECKSUMMED INLINE...")
WaitFor(self._get_checksum_record_status, test_file.name) \
.to_return_value('CHECKSUMMED', timeout_seconds=300)
# Verify that the inline checksum was not assigned a job id.
checksum_record = self._get_checksum_record(test_file.name)
self.assertIsNone(checksum_record.job_id)
# Check file record now contains checksums
db = self.db_session_maker.session()
file_record = db.query(DbFile).get(checksum_record.file_id)
self.assertEqual(test_file.checksums, file_record.checksums)
# Check S3 object has checksum tags
tagging = boto3.client('s3').get_object_tagging(
Bucket=self.upload_config.bucket_name,
Key=f"{self.upload_area_uuid}/{test_file.name}")
self.assertEqual(sorted(tagging['TagSet'], key=lambda x: x['Key']),
test_file.s3_tagset)
def _verify_file_is_checksummed_via_batch(self, test_file):
""" For files that are 10G or larger, we expect that the file will check-summed via batch. This means that it
first will need to be scheduled and the checksum record will be given a respective job id."""
print("VERIFYING FILE WAS CHECKSUMMED VIA BATCH...")
WaitFor(self._get_checksum_record_status, test_file.name) \
.to_return_value('SCHEDULED', timeout_seconds=30)
checksum_record = self._get_checksum_record(test_file.name)
WaitFor(self._get_batch_job_status, checksum_record.job_id) \
.to_return_value('SUCCEEDED', timeout_seconds=20 * MINUTE_SEC)
checksum_record = self._get_checksum_record(test_file.name)
self.assertEqual('CHECKSUMMED', checksum_record.status)
# Check file record now contains checksums
db = self.db_session_maker.session()
file_record = db.query(DbFile).get(checksum_record.file_id)
self.assertEqual(test_file.checksums, file_record.checksums)
# Check S3 object has checksum tags
tagging = boto3.client('s3').get_object_tagging(
Bucket=self.upload_config.bucket_name,
Key=f"{self.upload_area_uuid}/{test_file.name}")
self.assertEqual(sorted(tagging['TagSet'], key=lambda x: x['Key']),
test_file.s3_tagset)
def _verify_file_validation_status(self,
validation_id,
expected_exit_code=0,
expected_error_msg=''):
# Get the validation status of the file
_validation_results = self._get_validation_record(
validation_id).results
_actual_exit_code = _validation_results['exit_code']
_actual_error_msg = _validation_results['stdout']
self.assertEqual(expected_exit_code, _actual_exit_code)
self.assertIn(expected_error_msg, _actual_error_msg)
def _get_upload_area_record_status(self):
record = self._get_upload_area_record()
return record.status if record else None
def _get_checksum_record_status(self, filename):
record = self._get_checksum_record(filename)
return record.status if record else None
def _get_validation_record_job_id(self, validation_id):
record = self._get_validation_record(validation_id)
return record.job_id if record else None
def _get_validation_record_status(self, validation_id):
record = self._get_validation_record(validation_id)
return record.status if record else None
def _get_upload_area_record(self):
db = self.db_session_maker.session()
return db.query(DbUploadArea).filter(
DbUploadArea.uuid == self.upload_area_uuid).one_or_none()
def _get_checksum_record(self, filename):
db = self.db_session_maker.session()
s3_key = f"{self.upload_area_uuid}/{filename}"
file_record = db.query(DbFile).filter(
DbFile.s3_key == s3_key).one_or_none()
if file_record is None:
return None
checksum_record = db.query(DbChecksum).filter(
DbChecksum.file_id == file_record.id).one_or_none()
return checksum_record
def _get_validation_record(self, validation_id):
db = self.db_session_maker.session()
return db.query(DbValidation).filter(
DbValidation.id == validation_id).one_or_none()
def _get_batch_job_status(self, job_id):
response = self.batch.describe_jobs(jobs=[job_id])
self.assertEqual(1, len(response['jobs']))
return response['jobs'][0]['status']
def _make_request(self,
description,
verb,
url,
expected_status=None,
**options):
print(description + ": ")
print(f"{verb.upper()} {url}")
method = getattr(requests, verb.lower())
response = method(url, **options)
print(f"-> {response.status_code}")
if expected_status:
self.assertEqual(expected_status, response.status_code)
if response.content:
print(response.content.decode('utf8'))
return response.content
def _run_cli_command(self, description, command, expected_returncode=0):
print("\n" + description + ": ")
print(' '.join(command))
completed_process = subprocess.run(command, stdout=None, stderr=None)
self.assertEqual(expected_returncode, completed_process.returncode)
class TestUploadService(unittest.TestCase):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.batch = boto3.client('batch')
self.uri = None
self.db_session_maker = DBSessionMaker()
def setUp(self):
self.test_start_time = time.time()
self.upload_config = UploadConfig()
self.upload_client = UploadService(
deployment_stage=os.environ['DEPLOYMENT_STAGE'],
api_token=self.upload_config.api_key)
self.upload_area_uuid = "deadbeef-dead-dead-dead-%012d" % random.randint(
0, 999999999999)
print("")
self._execute_create_upload_area()
print("\tstartup time: %0.02f seconds." %
(time.time() - self.test_start_time))
def tearDown(self):
test_end_time = time.time()
print("\t%s took %0.02f seconds." %
(self._testMethodName, test_end_time - self.test_start_time))
self._execute_delete_upload_area()
print("\tteardown time: %0.02f seconds." %
(time.time() - test_end_time))
# All tests are formatted into 2-3 sections separated by blank lines:
#
# Setup preconditions (optional)
#
# Do the thing we are testing
#
# Test the thing was done
def test_store_file_using_api(self):
metadata_file = FixtureFile.factory('metadata_file.json')
self.upload_area.store_file(
filename=metadata_file.name,
file_content=metadata_file.contents,
content_type='application/json; dcp-type=metadata')
self._verify_file_was_checksummed_inline(
metadata_file) # Implicitly tests file was created.
def test_store_file_using_cli(self):
""" Tests storing of a file directly in S3, then notification of Upload via REST API """
small_file = FixtureFile.factory('small_file')
self._execute_upload_file_using_cli(small_file.path)
self._verify_file_was_checksummed_inline(
small_file) # Implicitly tests file was created.
def test_store_file_using_cli__with_large_file__triggers_batch_checksumming(
self):
large_file = FixtureFile.factory('10241MB_file')
self._execute_upload_file_using_cli(large_file.url)
self._verify_file_is_checksummed_via_batch(large_file)
def test_validate_file__with_valid_file__reports_validation_results(self):
small_file = FixtureFile.factory('small_file')
self.upload_area.store_file(
filename=small_file.name,
file_content=small_file.contents,
content_type='application/json; dcp-type=data')
response = self.upload_area.validate_files(
file_list=[small_file.name],
validator_image="humancellatlas/upload-validator-example:14")
validation_id = response['validation_id']
self._wait_for_validation_to_complete(validation_id)
self._verify_file_validation_status(
validation_id
) # default parameters checks for success in validation
def test__upload_invalid_file__validation_result_shows_invalid_state(self):
invalid_file = FixtureFile.factory('small_invalid_file')
self.upload_area.store_file(
filename=invalid_file.name,
file_content=invalid_file.contents,
content_type='application/json; dcp-type=data')
response = self.upload_area.validate_files(
file_list=[invalid_file.name],
validator_image="humancellatlas/upload-validator-example:14")
validation_id = response['validation_id']
self._wait_for_validation_to_complete(validation_id)
# Verify that the validation result of the file is invalid. This is designated by an exit code of 1 and the
# presence of an error message saying that file is invalid.
self._verify_file_validation_status(validation_id,
expected_exit_code=1,
expected_error_msg="invalid")
def _execute_create_upload_area(self):
self.upload_area = self.upload_client.create_area(
self.upload_area_uuid)
self.assertEqual('UNLOCKED', self._get_upload_area_record_status())
print(f"\tCreated upload area {self.upload_area_uuid}")
def _execute_upload_file_using_cli(self, file_location):
self._run_cli_command('hca', 'upload', 'select',
str(self.upload_area.uri))
self._run_cli_command('hca', 'upload', 'files', file_location)
self._run_cli_command('hca', 'upload', 'forget', self.upload_area.uuid)
def _wait_for_validation_to_complete(self, validation_id):
WaitFor(self._get_validation_record_status, validation_id) \
.to_return_value('SCHEDULED', timeout_seconds=MINUTE_SEC)
validation_job_id = self._get_validation_record_job_id(validation_id)
WaitFor(self._get_batch_job_status, validation_job_id) \
.to_return_value('SUCCEEDED', timeout_seconds=20 * MINUTE_SEC)
WaitFor(self._get_validation_record_status, validation_id) \
.to_return_value('VALIDATED', timeout_seconds=MINUTE_SEC)
def _execute_delete_upload_area(self):
print(f"\tDeleting upload area {self.upload_area.uuid}")
self.upload_area.delete()
WaitFor(self._get_upload_area_record_status) \
.to_return_value('DELETED', timeout_seconds=MINUTE_SEC)
def _verify_file_was_checksummed_inline(self, test_file):
""" For files that are smaller than 10G, we expect that the file will be check-summed inline. This means that
there is no need to schedule a job in batch and no job id is given to the checksum record."""
print("\tVerifying file was checksummed inline...")
WaitFor(self._get_checksum_record_status, test_file.name) \
.to_return_value('CHECKSUMMED', timeout_seconds=300)
# Verify that the inline checksum was not assigned a job id.
checksum_record = self._get_checksum_record(test_file.name)
self.assertIsNone(checksum_record.job_id)
# Check file record now contains checksums
db = self.db_session_maker.session()
file_record = db.query(DbFile).get(checksum_record.file_id)
self.assertEqual(test_file.checksums, file_record.checksums)
# Check S3 object has checksum tags
tagging = boto3.client('s3').get_object_tagging(
Bucket=self.upload_config.bucket_name,
Key=f"{self.upload_area_uuid}/{test_file.name}")
_actual_checksums = self._get_dict_representation_of_tagset_case_insensitive(
tagging['TagSet'])
_expected_checksums = self._get_dict_representation_of_tagset_case_insensitive(
test_file.s3_tagset)
self.assertDictEqual(_actual_checksums, _expected_checksums)
def _verify_file_is_checksummed_via_batch(self, test_file):
""" For files that are 10G or larger, we expect that the file will check-summed via batch. This means that it
first will need to be scheduled and the checksum record will be given a respective job id."""
print("\tVerifying file was checksummed via batch...")
WaitFor(self._get_checksum_record_status, test_file.name) \
.to_return_value('SCHEDULED', timeout_seconds=30)
checksum_record = self._get_checksum_record(test_file.name)
WaitFor(self._get_batch_job_status, checksum_record.job_id) \
.to_return_value('SUCCEEDED', timeout_seconds=20 * MINUTE_SEC)
checksum_record = self._get_checksum_record(test_file.name)
self.assertEqual('CHECKSUMMED', checksum_record.status)
# Check file record now contains checksums
db = self.db_session_maker.session()
file_record = db.query(DbFile).get(checksum_record.file_id)
[
self.assertEquals(
test_file.checksums[_checksum_function].lower(),
file_record.checksums[_checksum_function].lower())
for _checksum_function in set(
list(test_file.checksums.keys()) +
list(file_record.checksums.keys()))
]
# Check S3 object has checksum tags
tagging = boto3.client('s3').get_object_tagging(
Bucket=self.upload_config.bucket_name,
Key=f"{self.upload_area_uuid}/{test_file.name}")
_actual_checksums = self._get_dict_representation_of_tagset_case_insensitive(
tagging['TagSet'])
_expected_checksums = self._get_dict_representation_of_tagset_case_insensitive(
test_file.s3_tagset)
self.assertDictEqual(_actual_checksums, _expected_checksums)
def _verify_file_validation_status(self,
validation_id,
expected_exit_code=0,
expected_error_msg=''):
# Get the validation status of the file
_validation_results = self._get_validation_record(
validation_id).results
_actual_exit_code = _validation_results['exit_code']
_actual_error_msg = _validation_results['stdout']
self.assertEqual(expected_exit_code, _actual_exit_code)
self.assertIn(expected_error_msg, _actual_error_msg)
def _get_upload_area_record_status(self):
record = self._get_upload_area_record()
return record.status if record else None
def _get_checksum_record_status(self, filename):
record = self._get_checksum_record(filename)
return record.status if record else None
def _get_validation_record_job_id(self, validation_id):
record = self._get_validation_record(validation_id)
return record.job_id if record else None
def _get_validation_record_status(self, validation_id):
record = self._get_validation_record(validation_id)
return record.status if record else None
def _get_upload_area_record(self):
db = self.db_session_maker.session()
return db.query(DbUploadArea).filter(
DbUploadArea.uuid == self.upload_area_uuid).one_or_none()
def _get_checksum_record(self, filename):
db = self.db_session_maker.session()
s3_key = f"{self.upload_area_uuid}/{filename}"
file_record = db.query(DbFile).filter(
DbFile.s3_key == s3_key).one_or_none()
if file_record is None:
return None
checksum_record = db.query(DbChecksum).filter(
DbChecksum.file_id == file_record.id).one_or_none()
return checksum_record
def _get_validation_record(self, validation_id):
db = self.db_session_maker.session()
return db.query(DbValidation).filter(
DbValidation.id == validation_id).one_or_none()
def _get_batch_job_status(self, job_id):
response = self.batch.describe_jobs(jobs=[job_id])
self.assertEqual(1, len(response['jobs']))
return response['jobs'][0]['status']
def _run_cli_command(self, *command, expected_returncode=0):
print("\t" + ' '.join(command))
completed_process = subprocess.run(command, stdout=None, stderr=None)
self.assertEqual(expected_returncode, completed_process.returncode)
def _get_dict_representation_of_tagset_case_insensitive(self, tagset):
_tagset_dict = {}
for _item in tagset:
_tagset_dict[_item['Key'].lower()] = _item['Value'].lower()
return _tagset_dict