def _create_tables():
redshift = RedshiftHandler()
transaction = []
for table in TableName:
transaction.append(CREATE_QUERY_TEMPLATE[table.value].format("", "", table.value))
redshift.transaction(transaction)
def setUp(self):
self.dss_env = MATRIX_ENV_TO_DSS_ENV[os.environ['DEPLOYMENT_STAGE']]
self.api_url = f"https://{os.environ['API_HOST']}/v0"
self.res_dir = os.path.join(os.path.dirname(os.path.realpath(__file__)), "res")
self.headers = {'Content-type': 'application/json', 'Accept': 'application/json'}
self.verbose = True
self.s3_file_system = s3fs.S3FileSystem(anon=False)
self.redshift_handler = RedshiftHandler()
def __init__(self, bundle_uuid, bundle_version, event_type):
logger.info(
f"Running NotificationHandler with parameters: {bundle_uuid}, {bundle_version}, {event_type}"
)
self.bundle_uuid = bundle_uuid
self.bundle_version = bundle_version
self.event_type = event_type
self.redshift = RedshiftHandler()
def __init__(self, request_id: str, bundles_per_worker: int = 100):
Logging.set_correlation_id(logger, value=request_id)
self.request_id = request_id
self.bundles_per_worker = bundles_per_worker
self.request_tracker = RequestTracker(request_id)
self.dynamo_handler = DynamoHandler()
self.sqs_handler = SQSHandler()
self.infra_config = MatrixInfraConfig()
self.redshift_config = MatrixRedshiftConfig()
self.query_results_bucket = os.environ['MATRIX_QUERY_RESULTS_BUCKET']
self.s3_handler = S3Handler(os.environ['MATRIX_QUERY_BUCKET'])
self.redshift_handler = RedshiftHandler()
def load_tables(job_id: str, is_update: bool = False):
"""
Creates tables and loads PSVs from S3 into Redshift via SQL COPY.
:param job_id: UUID of the S3 prefix containinng the data to load in
:param is_update: True if existing database rows will be updated, else False
"""
_create_tables()
lock_query = """LOCK TABLE write_lock"""
delete_query_template = """DELETE FROM {0} USING {0}_temp WHERE {0}.{1} = {0}_temp.{1};"""
insert_query_template = """INSERT INTO {0} SELECT * FROM {0}_temp;"""
redshift = RedshiftHandler()
transaction = [lock_query]
for table in TableName:
if (is_update and table
== TableName.FEATURE) or table == TableName.WRITE_LOCK:
continue
s3_prefix = f"s3://{os.environ['MATRIX_PRELOAD_BUCKET']}/{job_id}/{table.value}"
if table == TableName.CELL:
s3_prefix += '/'
iam = os.environ['MATRIX_REDSHIFT_IAM_ROLE_ARN']
table_name = table.value if not is_update else f"{table.value}_temp"
if table == TableName.FEATURE:
copy_stmt = f"COPY {table_name} FROM '{s3_prefix}' iam_role '{iam}' COMPUPDATE ON;"
elif table == TableName.CELL:
copy_stmt = f"COPY {table_name} FROM '{s3_prefix}' iam_role '{iam}' GZIP COMPUPDATE ON;"
elif table == TableName.EXPRESSION:
copy_stmt = f"COPY {table_name} FROM '{s3_prefix}' iam_role '{iam}' GZIP COMPUPDATE ON COMPROWS 10000000;"
else:
copy_stmt = f"COPY {table_name} FROM '{s3_prefix}' iam_role '{iam}';"
if is_update:
logger.info(f"ETL: Building queries to update {table.value} table")
transaction.extend([
CREATE_QUERY_TEMPLATE[table.value].format(
"TEMP ", "_temp", table_name), copy_stmt,
delete_query_template.format(
table.value, RedshiftHandler.PRIMARY_KEY[table]),
insert_query_template.format(table.value)
])
else:
logger.info(f"ETL: Building queries to load {table_name} table")
transaction.append(copy_stmt)
logger.info(f"ETL: Populating Redshift tables. Committing transaction.")
try:
redshift.transaction(transaction)
except psycopg2.Error as e:
logger.error("Failed to populate Redshift tables. Rolling back.", e)
def _verify_load(es_query):
dss_client = etl.get_dss_client(deployment_stage=os.environ['DEPLOYMENT_STAGE'])
response = dss_client.post_search.iterate(es_query=es_query,
replica='aws',
per_page=500)
expected_bundles = list(result['bundle_fqid'] for result in response)
print(f"Loading {len(expected_bundles)} bundles to {os.environ['DEPLOYMENT_STAGE']} complete.\n"
f"Verifying row counts in Redshift...")
redshift = RedshiftHandler()
count_bundles_query = f"SELECT COUNT(*) FROM analysis WHERE bundle_fqid IN {format_str_list(expected_bundles)}"
results = redshift.transaction(queries=[count_bundles_query],
return_results=True)
print(f"Found {results[0][0]} analysis rows for {len(expected_bundles)} expected bundles.")
assert (results[0][0] == len(expected_bundles))
def _redshift_detail_lookup(name, description):
type_ = constants.METADATA_FIELD_TO_TYPE[name]
column_name = constants.METADATA_FIELD_TO_TABLE_COLUMN[name]
table_name = constants.TABLE_COLUMN_TO_TABLE[column_name]
fq_name = table_name + "." + column_name
table_primary_key = RedshiftHandler.PRIMARY_KEY[TableName(table_name)]
rs_handler = RedshiftHandler()
if type_ == "categorical":
query = query_constructor.create_field_detail_query(
fq_name, table_name, table_primary_key, type_)
results = dict(
rs_handler.transaction([query],
return_results=True,
read_only=True))
if None in results:
results[""] = results[None]
results.pop(None)
if True in results:
results["True"] = results[True]
results.pop(True)
if False in results:
results["False"] = results[False]
results.pop(False)
return ({
"field_name": name,
"field_description": description,
"field_type": type_,
"cell_counts": results
}, requests.codes.ok)
elif type_ == "numeric":
query = query_constructor.create_field_detail_query(
fq_name, table_name, table_primary_key, type_)
results = rs_handler.transaction([query],
return_results=True,
read_only=True)
min_ = results[0][0]
max_ = results[0][1]
return ({
"field_name": name,
"field_description": description,
"field_type": type_,
"minimum": min_,
"maximum": max_
}, requests.codes.ok)
#!/usr/bin/env python
"""
This script creates a readonly user in the redshift db
"""
from matrix.common.aws.redshift_handler import RedshiftHandler
from matrix.common.config import MatrixRedshiftConfig
matrix_config = MatrixRedshiftConfig()
redshift_handler = RedshiftHandler()
def handler():
readonly_username = matrix_config.readonly_username
readonly_password = matrix_config.readonly_password
drop_user_query = f"DROP USER IF EXISTS {readonly_username};"
add_user_query = f"CREATE USER {readonly_username} WITH PASSWORD '{readonly_password}';"
grant_public_query = f"grant select on all tables in schema public to {readonly_username};"
grant_information_schema_query = f"grant select on all tables in schema information_schema to {readonly_username};"
try:
redshift_handler.transaction([drop_user_query, add_user_query])
except Exception as e:
print(e)
redshift_handler.transaction(
[grant_public_query, grant_information_schema_query])
print("permissions applied to readonly user")
if __name__ == '__main__':
handler()
class TestMatrixServiceV0(MatrixServiceTest):
def setUp(self):
self.dss_env = MATRIX_ENV_TO_DSS_ENV[os.environ['DEPLOYMENT_STAGE']]
self.api_url = f"https://{os.environ['API_HOST']}/v0"
self.res_dir = os.path.join(
os.path.dirname(os.path.realpath(__file__)), "res")
self.headers = {
'Content-type': 'application/json',
'Accept': 'application/json'
}
self.verbose = True
self.s3_file_system = s3fs.S3FileSystem(anon=False)
self.redshift_handler = RedshiftHandler()
def tearDown(self):
if hasattr(self, 'request_id') and self.request_id:
self._cleanup_matrix_result(self.request_id)
def test_single_bundle_request(self):
self.request_id = self._post_matrix_service_request(
bundle_fqids=[INPUT_BUNDLE_IDS[self.dss_env][0]], format="loom")
WaitFor(self._poll_get_matrix_service_request, self.request_id)\
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=1200)
self._analyze_loom_matrix_results(self.request_id,
[INPUT_BUNDLE_IDS[self.dss_env][0]])
def test_loom_output_matrix_service(self):
self.request_id = self._post_matrix_service_request(
bundle_fqids=INPUT_BUNDLE_IDS[self.dss_env], format="loom")
# timeout seconds is increased to 1200 as batch may take time to spin up spot instances for conversion.
WaitFor(self._poll_get_matrix_service_request, self.request_id)\
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=1200)
self._analyze_loom_matrix_results(self.request_id,
INPUT_BUNDLE_IDS[self.dss_env])
def test_csv_output_matrix_service(self):
self.request_id = self._post_matrix_service_request(
bundle_fqids=INPUT_BUNDLE_IDS[self.dss_env], format="csv")
# timeout seconds is increased to 1200 as batch may take time to spin up spot instances for conversion.
WaitFor(self._poll_get_matrix_service_request, self.request_id) \
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=1200)
self._analyze_csv_matrix_results(self.request_id,
INPUT_BUNDLE_IDS[self.dss_env])
def test_mtx_output_matrix_service(self):
self.request_id = self._post_matrix_service_request(
bundle_fqids=INPUT_BUNDLE_IDS[self.dss_env], format="mtx")
# timeout seconds is increased to 1200 as batch may take time to spin up spot instances for conversion.
WaitFor(self._poll_get_matrix_service_request, self.request_id) \
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=1200)
self._analyze_mtx_matrix_results(self.request_id,
INPUT_BUNDLE_IDS[self.dss_env])
def test_cache_hit_matrix_service(self):
request_id_1 = self._post_matrix_service_request(
bundle_fqids=INPUT_BUNDLE_IDS[self.dss_env], format="loom")
# timeout seconds is increased to 1200 as batch may take time to spin up spot instances for conversion.
WaitFor(self._poll_get_matrix_service_request, request_id_1) \
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=1200)
self._analyze_loom_matrix_results(request_id_1,
INPUT_BUNDLE_IDS[self.dss_env])
request_id_2 = self._post_matrix_service_request(
bundle_fqids=INPUT_BUNDLE_IDS[self.dss_env], format="loom")
# timeout seconds is reduced to 300 as cache hits do not run conversion in batch.
WaitFor(self._poll_get_matrix_service_request, request_id_2) \
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=300)
self._analyze_loom_matrix_results(request_id_2,
INPUT_BUNDLE_IDS[self.dss_env])
matrix_location_1 = self._retrieve_matrix_location(request_id_1)
matrix_location_2 = self._retrieve_matrix_location(request_id_2)
loom_metrics_1 = validation.calculate_ss2_metrics_loom(
matrix_location_1)
loom_metrics_2 = validation.calculate_ss2_metrics_loom(
matrix_location_2)
self._compare_metrics(loom_metrics_1, loom_metrics_2)
self._cleanup_matrix_result(request_id_1)
self._cleanup_matrix_result(request_id_2)
def test_matrix_service_without_specified_output(self):
self.request_id = self._post_matrix_service_request(
bundle_fqids=INPUT_BUNDLE_IDS[self.dss_env])
WaitFor(self._poll_get_matrix_service_request, self.request_id)\
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=300)
self._analyze_loom_matrix_results(self.request_id,
INPUT_BUNDLE_IDS[self.dss_env])
def test_matrix_service_with_unexpected_bundles(self):
input_bundles = ['non-existent-bundle1', 'non-existent-bundle2']
self.request_id = self._post_matrix_service_request(
bundle_fqids=input_bundles)
WaitFor(self._poll_get_matrix_service_request, self.request_id)\
.to_return_value(MatrixRequestStatus.FAILED.value, timeout_seconds=300)
@unittest.skipUnless(
os.getenv('DEPLOYMENT_STAGE') != "prod",
"Do not want to process fake notifications in production.")
def test_dss_notification(self):
bundle_data = NOTIFICATION_TEST_DATA[self.dss_env]
bundle_fqid = bundle_data['bundle_fqid']
cell_row_count = bundle_data['cell_count']
expression_row_count = bundle_data['exp_count']
cellkeys = format_str_list(self._get_cellkeys_from_fqid(bundle_fqid))
self.assertTrue(len(cellkeys) > 0)
try:
self._post_notification(bundle_fqid=bundle_fqid,
event_type="DELETE")
WaitFor(self._poll_db_get_row_counts_for_fqid, bundle_fqid, cellkeys)\
.to_return_value((0, 0, 0), timeout_seconds=60)
self._post_notification(bundle_fqid=bundle_fqid,
event_type="CREATE")
WaitFor(self._poll_db_get_row_counts_for_fqid, bundle_fqid, cellkeys)\
.to_return_value((1, cell_row_count, expression_row_count), timeout_seconds=600)
self._post_notification(bundle_fqid=bundle_fqid,
event_type="TOMBSTONE")
WaitFor(self._poll_db_get_row_counts_for_fqid, bundle_fqid, cellkeys)\
.to_return_value((0, 0, 0), timeout_seconds=60)
self._post_notification(bundle_fqid=bundle_fqid,
event_type="UPDATE")
WaitFor(self._poll_db_get_row_counts_for_fqid, bundle_fqid, cellkeys)\
.to_return_value((1, cell_row_count, expression_row_count), timeout_seconds=600)
finally:
self._post_notification(bundle_fqid=bundle_fqid,
event_type="CREATE")
@unittest.skip
def test_matrix_service_ss2(self):
timeout = int(os.getenv("MATRIX_TEST_TIMEOUT", 300))
num_bundles = int(os.getenv("MATRIX_TEST_NUM_BUNDLES", 200))
bundle_fqids = json.loads(
open(f"{self.res_dir}/pancreas_ss2_2544_demo_bundles.json",
"r").read())[:num_bundles]
self.request_id = self._post_matrix_service_request(
bundle_fqids=bundle_fqids, format="loom")
# wait for request to complete
time.sleep(2)
WaitFor(self._poll_get_matrix_service_request, self.request_id)\
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=timeout)
self._analyze_loom_matrix_results(self.request_id, bundle_fqids)
def test_bundle_url(self):
timeout = int(os.getenv("MATRIX_TEST_TIMEOUT", 300))
bundle_fqids_url = INPUT_BUNDLE_URL.format(dss_env=self.dss_env)
self.request_id = self._post_matrix_service_request(
bundle_fqids_url=bundle_fqids_url, format="loom")
# wait for request to complete
WaitFor(self._poll_get_matrix_service_request, self.request_id)\
.to_return_value(MatrixRequestStatus.COMPLETE.value, timeout_seconds=timeout)
bundle_fqids = [
'.'.join(el.split('\t')) for el in requests.get(
bundle_fqids_url).text.strip().split('\n')[1:]
]
self._analyze_loom_matrix_results(self.request_id, bundle_fqids)
def _poll_db_get_row_counts_for_fqid(self, bundle_fqid, cellkeys):
"""
:return: Row counts associated with the given bundle fqid for (analysis_count, cell_count, exp_count)
"""
analysis_query = f"SELECT COUNT(*) FROM analysis WHERE analysis.bundle_fqid = '{bundle_fqid}'"
cell_query = f"SELECT COUNT(*) FROM cell WHERE cellkey IN {cellkeys}"
exp_query = f"SELECT COUNT(*) FROM expression WHERE cellkey IN {cellkeys}"
analysis_row_count = self.redshift_handler.transaction(
[analysis_query], return_results=True)[0][0]
cell_row_count = self.redshift_handler.transaction(
[cell_query], return_results=True)[0][0]
exp_row_count = self.redshift_handler.transaction(
[exp_query], return_results=True)[0][0]
return analysis_row_count, cell_row_count, exp_row_count
def _get_cellkeys_from_fqid(self, bundle_fqid):
"""
Returns a generator of cellkeys associated with a given bundle fqid.
"""
cellkeys_query = f"""
SELECT DISTINCT cellkey FROM cell
JOIN analysis ON cell.analysiskey = analysis.analysiskey
WHERE analysis.bundle_fqid = '{bundle_fqid}'
"""
results = self.redshift_handler.transaction([cellkeys_query],
return_results=True)
return (row[0] for row in results)
def _post_matrix_service_request(self,
bundle_fqids=None,
bundle_fqids_url=None,
format=None):
data = {}
if bundle_fqids:
data["bundle_fqids"] = bundle_fqids
if bundle_fqids_url:
data["bundle_fqids_url"] = bundle_fqids_url
if format:
data["format"] = format
response = self._make_request(
description="POST REQUEST TO MATRIX SERVICE",
verb='POST',
url=f"{self.api_url}/matrix",
expected_status=202,
data=json.dumps(data),
headers=self.headers)
data = json.loads(response)
return data["request_id"]
def _post_notification(self, bundle_fqid, event_type):
data = {}
bundle_uuid = bundle_fqid.split('.', 1)[0]
bundle_version = bundle_fqid.split('.', 1)[1]
data["transaction_id"] = "test_transaction_id"
data["subscription_id"] = "test_subscription_id"
data["event_type"] = event_type
data["match"] = {}
data["match"]["bundle_uuid"] = bundle_uuid
data["match"]["bundle_version"] = bundle_version
response = self._make_request(
description="POST NOTIFICATION TO MATRIX SERVICE",
verb='POST',
url=f"{self.api_url}/dss/notification",
expected_status=200,
data=json.dumps(data),
headers=self.headers)
data = json.loads(response)
return data
class NotificationHandler:
DELETE_ANALYSIS_QUERY_TEMPLATE = """DELETE FROM analysis WHERE bundle_fqid='{bundle_uuid}.{bundle_version}'"""
DELETE_CELL_QUERY_TEMPLATE = """
DELETE FROM cell
WHERE cell.analysiskey IN
(SELECT analysis.analysiskey FROM analysis
WHERE analysis.bundle_fqid='{bundle_uuid}.{bundle_version}')
"""
DELETE_EXPRESSION_QUERY_TEMPLATE = """
DELETE FROM expression
WHERE expression.cellkey IN
(SELECT cell.cellkey FROM cell
JOIN analysis ON cell.analysiskey = analysis.analysiskey
WHERE analysis.bundle_fqid='{bundle_uuid}.{bundle_version}')
"""
def __init__(self, bundle_uuid, bundle_version, event_type):
logger.info(
f"Running NotificationHandler with parameters: {bundle_uuid}, {bundle_version}, {event_type}"
)
self.bundle_uuid = bundle_uuid
self.bundle_version = bundle_version
self.event_type = event_type
self.redshift = RedshiftHandler()
def run(self):
if self.event_type == 'CREATE' or self.event_type == 'UPDATE':
self.update_bundle()
elif self.event_type == 'DELETE' or self.event_type == 'TOMBSTONE':
self.remove_bundle()
else:
logger.error(
f"Failed to process notification. Received invalid event type {self.event_type}."
)
return
logger.info(
f"Done processing DSS notification for {self.bundle_uuid}.{self.bundle_version}."
)
def update_bundle(self):
staging_dir = "/tmp"
content_type_patterns = ['application/json; dcp-type="metadata*"']
filename_patterns = [
"*zarr*", # match expression data
"*.results", # match SS2 results files
"*.mtx",
"genes.tsv",
"barcodes.tsv",
"empty_drops_result.csv"
] # match 10X results files
# clean up working directory in case of Lambda container reuse
shutil.rmtree(
f"{staging_dir}/{MetadataToPsvTransformer.OUTPUT_DIRNAME}",
ignore_errors=True)
etl.etl_dss_bundle(bundle_uuid=self.bundle_uuid,
bundle_version=self.bundle_version,
content_type_patterns=content_type_patterns,
filename_patterns=filename_patterns,
transformer_cb=etl.transform_bundle,
finalizer_cb=etl.finalizer_notification,
staging_directory=os.path.abspath(staging_dir),
deployment_stage=os.environ['DEPLOYMENT_STAGE'])
def remove_bundle(self):
self.redshift.transaction([
self.DELETE_EXPRESSION_QUERY_TEMPLATE.format(
bundle_uuid=self.bundle_uuid,
bundle_version=self.bundle_version),
self.DELETE_CELL_QUERY_TEMPLATE.format(
bundle_uuid=self.bundle_uuid,
bundle_version=self.bundle_version),
self.DELETE_ANALYSIS_QUERY_TEMPLATE.format(
bundle_uuid=self.bundle_uuid,
bundle_version=self.bundle_version)
])
class Driver:
"""
Formats and stores redshift queries in s3 and sqs for execution.
"""
def __init__(self, request_id: str, bundles_per_worker: int = 100):
Logging.set_correlation_id(logger, value=request_id)
self.request_id = request_id
self.bundles_per_worker = bundles_per_worker
self.request_tracker = RequestTracker(request_id)
self.dynamo_handler = DynamoHandler()
self.sqs_handler = SQSHandler()
self.infra_config = MatrixInfraConfig()
self.redshift_config = MatrixRedshiftConfig()
self.query_results_bucket = os.environ['MATRIX_QUERY_RESULTS_BUCKET']
self.s3_handler = S3Handler(os.environ['MATRIX_QUERY_BUCKET'])
self.redshift_handler = RedshiftHandler()
@property
def query_job_q_url(self):
return self.infra_config.query_job_q_url
@property
def redshift_role_arn(self):
return self.redshift_config.redshift_role_arn
def run(self, bundle_fqids: typing.List[str], bundle_fqids_url: str,
format: str):
"""
Initialize a matrix service request and spawn redshift queries.
:param bundle_fqids: List of bundle fqids to be queried on
:param bundle_fqids_url: URL from which bundle_fqids can be retrieved
:param format: MatrixFormat file format of output expression matrix
"""
logger.debug(
f"Driver running with parameters: bundle_fqids={bundle_fqids}, "
f"bundle_fqids_url={bundle_fqids_url}, format={format}, "
f"bundles_per_worker={self.bundles_per_worker}")
if bundle_fqids_url:
response = self._get_bundle_manifest(bundle_fqids_url)
resolved_bundle_fqids = self._parse_download_manifest(
response.text)
if len(resolved_bundle_fqids) == 0:
error_msg = "no bundles found in the supplied bundle manifest"
logger.info(error_msg)
self.request_tracker.log_error(error_msg)
return
else:
resolved_bundle_fqids = bundle_fqids
logger.debug(f"resolved bundles: {resolved_bundle_fqids}")
self.dynamo_handler.set_table_field_with_value(
DynamoTable.REQUEST_TABLE, self.request_id,
RequestTableField.NUM_BUNDLES, len(resolved_bundle_fqids))
s3_obj_keys = self._format_and_store_queries_in_s3(
resolved_bundle_fqids)
analysis_table_bundle_count = self._fetch_bundle_count_from_analysis_table(
resolved_bundle_fqids)
if analysis_table_bundle_count != len(resolved_bundle_fqids):
error_msg = "resolved bundles in request do not match bundles available in matrix service"
logger.info(error_msg)
self.request_tracker.log_error(error_msg)
return
for key in s3_obj_keys:
self._add_request_query_to_sqs(key, s3_obj_keys[key])
self.request_tracker.complete_subtask_execution(Subtask.DRIVER)
@retry(reraise=True, wait=wait_fixed(5), stop=stop_after_attempt(60))
def _get_bundle_manifest(self, bundle_fqids_url):
response = requests.get(bundle_fqids_url)
return response
@staticmethod
def _parse_download_manifest(data: str) -> typing.List[str]:
def _parse_line(line: str) -> str:
tokens = line.split("\t")
return f"{tokens[0]}.{tokens[1]}"
lines = data.splitlines()[1:]
return list(map(_parse_line, lines))
def _format_and_store_queries_in_s3(self, resolved_bundle_fqids: list):
feature_query = feature_query_template.format(
self.query_results_bucket, self.request_id, self.redshift_role_arn)
feature_query_obj_key = self.s3_handler.store_content_in_s3(
f"{self.request_id}/feature", feature_query)
exp_query = expression_query_template.format(
self.query_results_bucket, self.request_id, self.redshift_role_arn,
format_str_list(resolved_bundle_fqids))
exp_query_obj_key = self.s3_handler.store_content_in_s3(
f"{self.request_id}/expression", exp_query)
cell_query = cell_query_template.format(
self.query_results_bucket, self.request_id, self.redshift_role_arn,
format_str_list(resolved_bundle_fqids))
cell_query_obj_key = self.s3_handler.store_content_in_s3(
f"{self.request_id}/cell", cell_query)
return {
QueryType.CELL: cell_query_obj_key,
QueryType.EXPRESSION: exp_query_obj_key,
QueryType.FEATURE: feature_query_obj_key
}
def _add_request_query_to_sqs(self, query_type: QueryType,
s3_obj_key: str):
queue_url = self.query_job_q_url
payload = {
'request_id': self.request_id,
's3_obj_key': s3_obj_key,
'type': query_type.value
}
logger.debug(f"Adding {payload} to sqs {queue_url}")
self.sqs_handler.add_message_to_queue(queue_url, payload)
def _fetch_bundle_count_from_analysis_table(self,
resolved_bundle_fqids: list):
analysis_table_bundle_count_query = analysis_bundle_count_query_template.format(
format_str_list(resolved_bundle_fqids))
analysis_table_bundle_count_query = analysis_table_bundle_count_query.strip(
).replace('\n', '')
results = self.redshift_handler.transaction(
[analysis_table_bundle_count_query],
read_only=True,
return_results=True)
analysis_table_bundle_count = results[0][0]
return analysis_table_bundle_count
def __init__(self):
self.sqs_handler = SQSHandler()
self.s3_handler = S3Handler(os.environ["MATRIX_QUERY_BUCKET"])
self.batch_handler = BatchHandler()
self.redshift_handler = RedshiftHandler()
self.matrix_infra_config = MatrixInfraConfig()
class QueryRunner:
def __init__(self):
self.sqs_handler = SQSHandler()
self.s3_handler = S3Handler(os.environ["MATRIX_QUERY_BUCKET"])
self.batch_handler = BatchHandler()
self.redshift_handler = RedshiftHandler()
self.matrix_infra_config = MatrixInfraConfig()
@property
def query_job_q_url(self):
return self.matrix_infra_config.query_job_q_url
@property
def query_job_deadletter_q_url(self):
return self.matrix_infra_config.query_job_deadletter_q_url
def run(self, max_loops=None):
loops = 0
while max_loops is None or loops < max_loops:
loops += 1
messages = self.sqs_handler.receive_messages_from_queue(
self.query_job_q_url)
if messages:
message = messages[0]
logger.info(f"Received {message} from {self.query_job_q_url}")
payload = json.loads(message['Body'])
request_id = payload['request_id']
request_tracker = RequestTracker(request_id)
Logging.set_correlation_id(logger, value=request_id)
obj_key = payload['s3_obj_key']
receipt_handle = message['ReceiptHandle']
try:
logger.info(f"Fetching query from {obj_key}")
query = self.s3_handler.load_content_from_obj_key(obj_key)
logger.info(f"Running query from {obj_key}")
self.redshift_handler.transaction([query], read_only=True)
logger.info(f"Finished running query from {obj_key}")
logger.info(
f"Deleting {message} from {self.query_job_q_url}")
self.sqs_handler.delete_message_from_queue(
self.query_job_q_url, receipt_handle)
logger.info(
"Incrementing completed queries in state table")
request_tracker.complete_subtask_execution(Subtask.QUERY)
if request_tracker.is_request_ready_for_conversion():
logger.info("Scheduling batch conversion job")
batch_job_id = self.batch_handler.schedule_matrix_conversion(
request_id, request_tracker.format)
request_tracker.write_batch_job_id_to_db(batch_job_id)
except Exception as e:
logger.info(
f"QueryRunner failed on {message} with error {e}")
request_tracker.log_error(str(e))
logger.info(
f"Adding {message} to {self.query_job_deadletter_q_url}"
)
self.sqs_handler.add_message_to_queue(
self.query_job_deadletter_q_url, payload)
logger.info(
f"Deleting {message} from {self.query_job_q_url}")
self.sqs_handler.delete_message_from_queue(
self.query_job_q_url, receipt_handle)
else:
logger.info(f"No messages to read from {self.query_job_q_url}")
