populator.max_row_count = args.max_row_count
populator.allowed_storage_formats = args.storage_file_formats.split(
',')
if args.use_hive:
with cluster.hive.connect() as conn:
with conn.cursor() as cursor:
cursor.ensure_empty_db(args.db_name)
else:
with cluster.impala.connect() as conn:
with conn.cursor() as cursor:
cursor.invalidate_metadata()
cursor.ensure_empty_db(args.db_name)
if args.use_postgresql:
with cli_options.create_connection(args) as postgresql_conn:
with postgresql_conn.cursor() as cursor:
cursor.ensure_empty_db(args.db_name)
postgresql_conn = cli_options.create_connection(
args, db_name=args.db_name)
else:
postgresql_conn = None
populator.populate_db(args.table_count,
postgresql_conn=postgresql_conn)
else:
if args.migrate_table_names:
table_names = args.migrate_table_names.split(',')
else:
table_names = None
with cli_options.create_connection(args) as conn:
with conn.cursor() as cursor:
for profile in PROFILES:
profile_name = profile.__name__
if profile_name.endswith('Profile'):
profile_name = profile_name[:-1 * len('Profile')]
profiles[profile_name.lower()] = profile
parser.add_argument('--profile', default='default',
choices=(sorted(profiles.keys())),
help='Determines the mix of SQL features to use during query generation.')
# TODO: Seed the random query generator for repeatable queries?
args = parser.parse_args()
cli_options.configure_logging(
args.log_level, debug_log_file=args.debug_log_file, log_thread_name=True)
cluster = cli_options.create_cluster(args)
ref_conn = cli_options.create_connection(args, args.ref_db_type, db_name=args.db_name)
if args.test_db_type == IMPALA:
test_conn = cluster.impala.connect(db_name=args.db_name)
elif args.test_db_type == HIVE:
test_conn = cluster.hive.connect(db_name=args.db_name)
else:
test_conn = cli_options.create_connection(
args, args.test_db_type, db_name=args.db_name)
# Create an instance of profile class (e.g. DefaultProfile)
query_profile = profiles[args.profile]()
if args.explain_only:
searcher = FrontendExceptionSearcher(query_profile, ref_conn, test_conn)
searcher.search(args.query_count)
else:
diff_searcher = QueryResultDiffSearcher(query_profile, ref_conn, test_conn)
query_timeout_seconds = args.timeout
parser.add_argument(
'--profile',
default='default',
choices=(sorted(profiles.keys())),
help=
'Determines the mix of SQL features to use during query generation.')
# TODO: Seed the random query generator for repeatable queries?
args = parser.parse_args()
cli_options.configure_logging(args.log_level,
debug_log_file=args.debug_log_file,
log_thread_name=True)
cluster = cli_options.create_cluster(args)
ref_conn = cli_options.create_connection(args,
args.ref_db_type,
db_name=args.db_name)
if args.test_db_type == IMPALA:
test_conn = cluster.impala.connect(db_name=args.db_name)
elif args.test_db_type == HIVE:
test_conn = cluster.hive.connect(db_name=args.db_name)
else:
test_conn = cli_options.create_connection(args,
args.test_db_type,
db_name=args.db_name)
# Create an instance of profile class (e.g. DefaultProfile)
query_profile = profiles[args.profile]()
if args.explain_only:
searcher = FrontendExceptionSearcher(query_profile, ref_conn,
test_conn)
searcher.search(args.query_count)
populator.min_row_count = args.min_row_count
populator.max_row_count = args.max_row_count
populator.allowed_storage_formats = args.storage_file_formats.split(',')
if args.use_hive:
with cluster.hive.connect() as conn:
with conn.cursor() as cursor:
cursor.ensure_empty_db(args.db_name)
else:
with cluster.impala.connect() as conn:
with conn.cursor() as cursor:
cursor.invalidate_metadata()
cursor.ensure_empty_db(args.db_name)
if args.use_postgresql:
with cli_options.create_connection(args) as postgresql_conn:
with postgresql_conn.cursor() as cursor:
cursor.ensure_empty_db(args.db_name)
postgresql_conn = cli_options.create_connection(args, db_name=args.db_name)
else:
postgresql_conn = None
populator.populate_db(args.table_count, postgresql_conn=postgresql_conn)
else:
if args.migrate_table_names:
table_names = args.migrate_table_names.split(',')
else:
table_names = None
with cli_options.create_connection(args) as conn:
with conn.cursor() as cursor:
cursor.ensure_empty_db(args.db_name)
with cli_options.create_connection(args, db_name=args.db_name) as conn: