序
本文主要研究一下clickhouse分布式表的操作
创建分布式表
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
(name1 [type1] [DEFAULT|MATERIALIZED|ALIAS expr1],name2 [type2] [DEFAULT|MATERIALIZED|ALIAS expr2],...
) ENGINE = Distributed(cluster, database, table[, sharding_key[, policy_name]])
[SETTINGS name=value, ...]
示例
CREATE TABLE distribute_test_all on cluster c1
(name String,age Int16,weight Float32
)
ENGINE = Distributed(c1, db1, distribute_test_local, rand());
这里在c1的cluster定义了一个distribute_test_all的分布式表，查询本地的distribute_test_local表，而sharding_key采用的是随机的方式，将数据分布到每个节点的本地表
创建本地表
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
(name1 [type1] [NULL|NOT NULL] [DEFAULT|MATERIALIZED|EPHEMERAL|ALIAS expr1] [compression_codec] [TTL expr1] [COMMENT 'comment for column'],name2 [type2] [NULL|NOT NULL] [DEFAULT|MATERIALIZED|EPHEMERAL|ALIAS expr2] [compression_codec] [TTL expr2] [COMMENT 'comment for column'],...
) ENGINE = engineCOMMENT 'comment for table'
示例
CREATE TABLE distribute_test_local on cluster c1
(name String,age Int16,weight Float32
)
Engine=MergeTree()
ORDER BY name;
这里在c1的cluster的每个实例上创建distribute_test_local表，engine为MergeTree
删除表
--删除分布式表
DROP TABLE distribute_test_all ON CLUSTER c1-删除本地表
DROP TABLE distribute_test_local ON CLUSTER c1
如果要彻底删除一张分布式表则需要分别删除分布式表和本地表
小结
clickhouse提供了分布式表，针对insert和select的查询，会以分布式的方式作用于本地表。
