| 语法 | CREATE/ALTER/DROP … ON CLUSTER <cluster> |
|---|---|
| 查看进度 | SELECT * FROM system.distributed_ddl_queue 按 query_id/状态过滤 |
| 注意事项 | 各节点结构必须一致，否则副本表会出现同步异常 |
示例
示例
CREATE DATABASE IF NOT EXISTS analytics ON CLUSTER prod;
CREATE TABLE analytics.events_local ON CLUSTER prod
(
ts DateTime,
id UInt64
)
ENGINE = ReplicatedMergeTree('/clickhouse/tables/{shard}/events_local', '{replica}')
ORDER BY (id, ts);
SELECT entry, status, status_stopped_host
FROM system.distributed_ddl_queue
ORDER BY entry DESC LIMIT 10;分布式 DDL 需要一个健康的协调服务与同名集群配置；部分节点失败时可用 SYSTEM RESTART REPLICA 或重跑语句处理。