mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-11-17 21:24:28 +00:00
95 lines
2.7 KiB
Python
95 lines
2.7 KiB
Python
import pytest
|
|
from helpers.cluster import ClickHouseCluster
|
|
|
|
cluster = ClickHouseCluster(__file__)
|
|
|
|
nodes = [
|
|
cluster.add_instance(
|
|
f"n{i}", main_configs=["configs/remote_servers.xml"], with_zookeeper=True
|
|
)
|
|
for i in range(1, 5)
|
|
]
|
|
|
|
|
|
@pytest.fixture(scope="module", autouse=True)
|
|
def start_cluster():
|
|
try:
|
|
cluster.start()
|
|
yield cluster
|
|
finally:
|
|
cluster.shutdown()
|
|
|
|
|
|
def create_tables(cluster):
|
|
n1 = nodes[0]
|
|
n1.query("DROP TABLE IF EXISTS dist_table")
|
|
n1.query(f"DROP TABLE IF EXISTS test_table ON CLUSTER {cluster}")
|
|
|
|
n1.query(
|
|
f"CREATE TABLE test_table ON CLUSTER {cluster} (key Int32, value String) Engine=MergeTree ORDER BY (key, sipHash64(value))"
|
|
)
|
|
n1.query(
|
|
f"""
|
|
CREATE TABLE dist_table AS test_table
|
|
Engine=Distributed(
|
|
{cluster},
|
|
currentDatabase(),
|
|
test_table,
|
|
rand()
|
|
)
|
|
"""
|
|
)
|
|
|
|
|
|
def insert_data(cluster, row_num):
|
|
create_tables(cluster)
|
|
n1 = nodes[0]
|
|
n1.query(
|
|
f"INSERT INTO dist_table SELECT number % 4, number FROM numbers({row_num})"
|
|
)
|
|
n1.query("SYSTEM FLUSH DISTRIBUTED dist_table")
|
|
|
|
|
|
@pytest.mark.parametrize("custom_key", ["sipHash64(key)", "key"])
|
|
@pytest.mark.parametrize("filter_type", ["default", "range"])
|
|
@pytest.mark.parametrize(
|
|
"cluster",
|
|
["test_multiple_shards_multiple_replicas", "test_single_shard_multiple_replicas"],
|
|
)
|
|
def test_parallel_replicas_custom_key(start_cluster, cluster, custom_key, filter_type):
|
|
for node in nodes:
|
|
node.rotate_logs()
|
|
|
|
row_num = 1000
|
|
insert_data(cluster, row_num)
|
|
|
|
expected_result = ""
|
|
for i in range(4):
|
|
expected_result += f"{i}\t250\n"
|
|
|
|
n1 = nodes[0]
|
|
assert (
|
|
n1.query(
|
|
"SELECT key, count() FROM dist_table GROUP BY key ORDER BY key",
|
|
settings={
|
|
"prefer_localhost_replica": 0,
|
|
"max_parallel_replicas": 4,
|
|
"parallel_replicas_custom_key": custom_key,
|
|
"parallel_replicas_custom_key_filter_type": filter_type,
|
|
},
|
|
)
|
|
== expected_result
|
|
)
|
|
|
|
if cluster == "test_multiple_shards_multiple_replicas":
|
|
# we simply process query on all replicas for each shard by appending the filter on replica
|
|
assert all(
|
|
node.contains_in_log("Processing query on a replica using custom_key")
|
|
for node in nodes
|
|
)
|
|
else:
|
|
# we first transform all replicas into shards and then append for each shard filter
|
|
assert n1.contains_in_log(
|
|
"Single shard cluster used with custom_key, transforming replicas into virtual shards"
|
|
)
|