How Do I Delete the Backup Table of the ZooKeeper Node in the ClickHouse Cluster?
Symptom
When you connect a ClickHouse cluster and create a table, the ZooKeeper node does not get fully removed after you delete the table.
Figure 1 Execution results

Possible Causes
The command for creating a table contains ON CLUSTER default_cluster. Therefore, a local table is created on each node.
Solutions
Delete the tables and stored data on other nodes.
- Create a database.create database demo ON CLUSTER default_cluster;
- Create tables in the database.use demo;
Create the test table.
CREATE TABLE demo.test ON CLUSTER default_cluster(`EventDate` DateTime, `id` UInt64)ENGINE = ReplicatedMergeTree('/clickhouse/tables/{shard}/default/test', '{replica}') PARTITION BY toYYYYMM(EventDate) ORDER BY id; - Delete the table.drop table test SYNC;
- The following error message is displayed if a table with the same name as the deleted one is created:┌─host──────────┬─port─┬─status─┬─error──────────────────────────────────────────────────────────────────────────────────────────────┬─num_hosts_remaining─┬─num_hosts_active─┐│ 192.168.2.185 │ 9000 │ 57 │ Code: 57. DB::Exception: Table demo.test already exists. (TABLE_ALREADY_EXISTS) (version 22.3.2.1) │ 1 │ 1 │└───────────────┴──────┴────────┴────────────────────────────────────────────────────────────────────────────────────────────────────┴─────────────────────┴──────────────────┘┌─host─────────┬─port─┬─status─┬─error─┬─num_hosts_remaining─┬─num_hosts_active─┐│ 192.168.2.16 │ 9000 │ 0 │ │ 0 │ 0 │└──────────────┴──────┴────────┴───────┴─────────────────────┴──────────────────┘
- Method 1: Run the following command to delete the tables and related data on each node:drop table test ON CLUSTER default_cluster SYNC;
No error occurs if a table with the same name as the deleted one is created.
- Method 2: Delete the table on other nodes.drop table test SYNC;
Parent topic: Data Read/Write
- Symptom
- Possible Causes
- Solutions