ClickHouse/tests/integration/test_ddl_alter_query/test.py
bharatnc b75963d370 Format and optimize imports in integration test files
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.
2020-09-15 23:10:54 -07:00

52 lines
2.3 KiB
Python

import pytest
from helpers.cluster import ClickHouseCluster
cluster = ClickHouseCluster(__file__)
node1 = cluster.add_instance('node1', main_configs=['configs/remote_servers.xml'], with_zookeeper=True)
node2 = cluster.add_instance('node2', main_configs=['configs/remote_servers.xml'], with_zookeeper=True)
node3 = cluster.add_instance('node3', main_configs=['configs/remote_servers.xml'], with_zookeeper=True)
node4 = cluster.add_instance('node4', main_configs=['configs/remote_servers.xml'], with_zookeeper=True)
@pytest.fixture(scope="module")
def started_cluster():
try:
cluster.start()
for i, node in enumerate([node1, node2]):
node.query("CREATE DATABASE testdb")
node.query(
'''CREATE TABLE testdb.test_table(id UInt32, val String) ENGINE = ReplicatedMergeTree('/clickhouse/test/test_table1', '{}') ORDER BY id;'''.format(
i))
for i, node in enumerate([node3, node4]):
node.query("CREATE DATABASE testdb")
node.query(
'''CREATE TABLE testdb.test_table(id UInt32, val String) ENGINE = ReplicatedMergeTree('/clickhouse/test/test_table2', '{}') ORDER BY id;'''.format(
i))
yield cluster
finally:
cluster.shutdown()
def test_alter(started_cluster):
node1.query("INSERT INTO testdb.test_table SELECT number, toString(number) FROM numbers(100)")
node3.query("INSERT INTO testdb.test_table SELECT number, toString(number) FROM numbers(100)")
node2.query("SYSTEM SYNC REPLICA testdb.test_table")
node4.query("SYSTEM SYNC REPLICA testdb.test_table")
node1.query("ALTER TABLE testdb.test_table ON CLUSTER test_cluster ADD COLUMN somecolumn UInt8 AFTER val",
settings={"replication_alter_partitions_sync": "2"})
node1.query("SYSTEM SYNC REPLICA testdb.test_table")
node2.query("SYSTEM SYNC REPLICA testdb.test_table")
node3.query("SYSTEM SYNC REPLICA testdb.test_table")
node4.query("SYSTEM SYNC REPLICA testdb.test_table")
assert node1.query("SELECT somecolumn FROM testdb.test_table LIMIT 1") == "0\n"
assert node2.query("SELECT somecolumn FROM testdb.test_table LIMIT 1") == "0\n"
assert node3.query("SELECT somecolumn FROM testdb.test_table LIMIT 1") == "0\n"
assert node4.query("SELECT somecolumn FROM testdb.test_table LIMIT 1") == "0\n"