mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-11-10 17:44:23 +00:00
169 lines
5.7 KiB
Python
169 lines
5.7 KiB
Python
import pytest
|
|
import threading
|
|
import time
|
|
from helpers.cluster import ClickHouseCluster
|
|
|
|
cluster = ClickHouseCluster(__file__)
|
|
|
|
node1 = cluster.add_instance('node1',
|
|
config_dir='configs',
|
|
main_configs=['configs/logs_config.xml'],
|
|
with_zookeeper=True,
|
|
macros={"shard": 0, "replica": 1} )
|
|
|
|
node2 = cluster.add_instance('node2',
|
|
config_dir='configs',
|
|
main_configs=['configs/logs_config.xml'],
|
|
with_zookeeper=True,
|
|
macros={"shard": 0, "replica": 2} )
|
|
|
|
|
|
@pytest.fixture(scope="module")
|
|
def started_cluster():
|
|
try:
|
|
cluster.start()
|
|
node1.query('CREATE DATABASE test ENGINE=Ordinary')
|
|
node2.query('CREATE DATABASE test ENGINE=Ordinary')
|
|
yield cluster
|
|
|
|
finally:
|
|
cluster.shutdown()
|
|
|
|
|
|
def split_tsv(data):
|
|
return [ x.split("\t") for x in data.splitlines() ]
|
|
|
|
|
|
@pytest.mark.parametrize("replicated", [
|
|
"",
|
|
"replicated"
|
|
])
|
|
def test_merge_simple(started_cluster, replicated):
|
|
try:
|
|
clickhouse_path = "/var/lib/clickhouse"
|
|
db_name = "test"
|
|
table_name = "merge_simple"
|
|
name = db_name + "." + table_name
|
|
table_path = "data/" + db_name + "/" + table_name
|
|
nodes = [node1, node2] if replicated else [node1]
|
|
engine = "ReplicatedMergeTree('/clickhouse/test_merge_simple', '{replica}')" if replicated else "MergeTree()"
|
|
node_check = nodes[-1]
|
|
starting_block = 0 if replicated else 1
|
|
|
|
for node in nodes:
|
|
node.query("""
|
|
CREATE TABLE {name}
|
|
(
|
|
`a` Int64
|
|
)
|
|
ENGINE = {engine}
|
|
ORDER BY sleep(2)
|
|
""".format(engine=engine, name=name))
|
|
|
|
node1.query("INSERT INTO {name} VALUES (1)".format(name=name))
|
|
node1.query("INSERT INTO {name} VALUES (2)".format(name=name))
|
|
node1.query("INSERT INTO {name} VALUES (3)".format(name=name))
|
|
|
|
parts = ["all_{}_{}_0".format(x, x) for x in range(starting_block, starting_block+3)]
|
|
result_part = "all_{}_{}_1".format(starting_block, starting_block+2)
|
|
|
|
def optimize():
|
|
node1.query("OPTIMIZE TABLE {name}".format(name=name))
|
|
|
|
wait = threading.Thread(target=time.sleep, args=(5,))
|
|
wait.start()
|
|
t = threading.Thread(target=optimize)
|
|
t.start()
|
|
|
|
time.sleep(1)
|
|
assert split_tsv(node_check.query("""
|
|
SELECT database, table, num_parts, source_part_names, source_part_paths, result_part_name, result_part_path, partition_id, is_mutation
|
|
FROM system.merges
|
|
WHERE table = '{name}'
|
|
""".format(name=table_name))) == [
|
|
[
|
|
db_name,
|
|
table_name,
|
|
"3",
|
|
"['{}','{}','{}']".format(*parts),
|
|
"['{clickhouse}/{table_path}/{}/','{clickhouse}/{table_path}/{}/','{clickhouse}/{table_path}/{}/']".format(*parts, clickhouse=clickhouse_path, table_path=table_path),
|
|
result_part,
|
|
"{clickhouse}/{table_path}/{}/".format(result_part, clickhouse=clickhouse_path, table_path=table_path),
|
|
"all",
|
|
"0"
|
|
]
|
|
]
|
|
t.join()
|
|
wait.join()
|
|
|
|
assert node_check.query("SELECT * FROM system.merges WHERE table = '{name}'".format(name=table_name)) == ""
|
|
|
|
finally:
|
|
for node in nodes:
|
|
node.query("DROP TABLE {name}".format(name=name))
|
|
|
|
|
|
@pytest.mark.parametrize("replicated", [
|
|
"",
|
|
"replicated"
|
|
])
|
|
def test_mutation_simple(started_cluster, replicated):
|
|
try:
|
|
clickhouse_path = "/var/lib/clickhouse"
|
|
db_name = "test"
|
|
table_name = "mutation_simple"
|
|
name = db_name + "." + table_name
|
|
table_path = "data/" + db_name + "/" + table_name
|
|
nodes = [node1, node2] if replicated else [node1]
|
|
engine = "ReplicatedMergeTree('/clickhouse/test_mutation_simple', '{replica}')" if replicated else "MergeTree()"
|
|
node_check = nodes[-1]
|
|
starting_block = 0 if replicated else 1
|
|
|
|
for node in nodes:
|
|
node.query("""
|
|
CREATE TABLE {name}
|
|
(
|
|
`a` Int64
|
|
)
|
|
ENGINE = {engine}
|
|
ORDER BY tuple()
|
|
""".format(engine=engine, name=name))
|
|
|
|
node1.query("INSERT INTO {name} VALUES (1)".format(name=name))
|
|
part = "all_{}_{}_0".format(starting_block, starting_block)
|
|
result_part = "all_{}_{}_0_{}".format(starting_block, starting_block, starting_block+1)
|
|
|
|
def alter():
|
|
node1.query("ALTER TABLE {name} UPDATE a = 42 WHERE sleep(2) OR 1".format(name=name))
|
|
|
|
t = threading.Thread(target=alter)
|
|
t.start()
|
|
|
|
time.sleep(1)
|
|
assert split_tsv(node_check.query("""
|
|
SELECT database, table, num_parts, source_part_names, source_part_paths, result_part_name, result_part_path, partition_id, is_mutation
|
|
FROM system.merges
|
|
WHERE table = '{name}'
|
|
""".format(name=table_name))) == [
|
|
[
|
|
db_name,
|
|
table_name,
|
|
"1",
|
|
"['{}']".format(part),
|
|
"['{clickhouse}/{table_path}/{}/']".format(part, clickhouse=clickhouse_path, table_path=table_path),
|
|
result_part,
|
|
"{clickhouse}/{table_path}/{}/".format(result_part, clickhouse=clickhouse_path, table_path=table_path),
|
|
"all",
|
|
"1"
|
|
],
|
|
]
|
|
t.join()
|
|
|
|
time.sleep(1.5)
|
|
|
|
assert node_check.query("SELECT * FROM system.merges WHERE table = '{name}'".format(name=table_name)) == ""
|
|
|
|
finally:
|
|
for node in nodes:
|
|
node.query("DROP TABLE {name}".format(name=name))
|