mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-11-22 07:31:57 +00:00
b75963d370
This PR formats all the `*.py` files found under the `tests/integration` folder. It also reorders the imports and cleans up a bunch of unused imports. The formatting also takes care of other things like wrapping lines and fixing spaces and indents such that the tests look more readable.
77 lines
3.3 KiB
Python
77 lines
3.3 KiB
Python
import pytest
|
|
|
|
from helpers.cluster import ClickHouseCluster
|
|
from helpers.test_tools import assert_eq_with_retry
|
|
|
|
cluster = ClickHouseCluster(__file__)
|
|
|
|
node1 = cluster.add_instance('node1', main_configs=['configs/remote_servers.xml'],
|
|
user_configs=['configs/user_good_restricted.xml'], with_zookeeper=True)
|
|
node2 = cluster.add_instance('node2', main_configs=['configs/remote_servers.xml'],
|
|
user_configs=['configs/user_good_restricted.xml'], with_zookeeper=True)
|
|
node3 = cluster.add_instance('node3', main_configs=['configs/remote_servers.xml'],
|
|
user_configs=['configs/user_good_allowed.xml'], with_zookeeper=True)
|
|
node4 = cluster.add_instance('node4', main_configs=['configs/remote_servers.xml'],
|
|
user_configs=['configs/user_good_allowed.xml'], with_zookeeper=True)
|
|
|
|
|
|
@pytest.fixture(scope="module")
|
|
def started_cluster():
|
|
try:
|
|
cluster.start()
|
|
|
|
for node in [node1, node2]:
|
|
node.query('''
|
|
CREATE TABLE sometable(date Date, id UInt32, value Int32)
|
|
ENGINE = ReplicatedMergeTree('/clickhouse/tables/0/sometable', '{replica}', date, id, 8192);
|
|
'''.format(replica=node.name), user='awesome')
|
|
|
|
for node in [node3, node4]:
|
|
node.query('''
|
|
CREATE TABLE someothertable(date Date, id UInt32, value Int32)
|
|
ENGINE = ReplicatedMergeTree('/clickhouse/tables/0/someothertable', '{replica}', date, id, 8192);
|
|
'''.format(replica=node.name), user='good')
|
|
|
|
yield cluster
|
|
|
|
finally:
|
|
cluster.shutdown()
|
|
|
|
|
|
@pytest.mark.parametrize("table,query,expected,n1,n2", [
|
|
("sometable", "ALTER TABLE sometable DROP PARTITION 201706", '1', node1, node2),
|
|
("sometable", "TRUNCATE TABLE sometable", '0', node1, node2),
|
|
("sometable", "OPTIMIZE TABLE sometable", '4', node1, node2),
|
|
("someothertable", "ALTER TABLE someothertable DROP PARTITION 201706", '1', node3, node4),
|
|
("someothertable", "TRUNCATE TABLE someothertable", '0', node3, node4),
|
|
("someothertable", "OPTIMIZE TABLE someothertable", '4', node3, node4),
|
|
])
|
|
def test_alter_table_drop_partition(started_cluster, table, query, expected, n1, n2):
|
|
to_insert = '''\
|
|
2017-06-16 111 0
|
|
2017-06-16 222 1
|
|
2017-06-16 333 2
|
|
2017-07-16 444 3
|
|
'''
|
|
n1.query("INSERT INTO {} FORMAT TSV".format(table), stdin=to_insert, user='good')
|
|
|
|
assert_eq_with_retry(n1, "SELECT COUNT(*) from {}".format(table), '4', user='good')
|
|
assert_eq_with_retry(n2, "SELECT COUNT(*) from {}".format(table), '4', user='good')
|
|
|
|
### It maybe leader and everything will be ok
|
|
n1.query(query, user='good')
|
|
|
|
assert_eq_with_retry(n1, "SELECT COUNT(*) from {}".format(table), expected, user='good')
|
|
assert_eq_with_retry(n2, "SELECT COUNT(*) from {}".format(table), expected, user='good')
|
|
|
|
n1.query("INSERT INTO {} FORMAT TSV".format(table), stdin=to_insert, user='good')
|
|
|
|
assert_eq_with_retry(n1, "SELECT COUNT(*) from {}".format(table), '4', user='good')
|
|
assert_eq_with_retry(n2, "SELECT COUNT(*) from {}".format(table), '4', user='good')
|
|
|
|
### If node1 is leader than node2 will be slave
|
|
n2.query(query, user='good')
|
|
|
|
assert_eq_with_retry(n1, "SELECT COUNT(*) from {}".format(table), expected, user='good')
|
|
assert_eq_with_retry(n2, "SELECT COUNT(*) from {}".format(table), expected, user='good')
|