How was the configuration of ClickHouse set up in shards and replicas?

In ClickHouse configuration for sharding and replication, settings are usually used in the config.xml and users.xml files. For shards, separate clusters are specified with server addresses, and for replicas, ZooKeeper is configured to coordinate and synchronize data between replicas.

An example configuration of a cluster with two shard servers and replicas:

<remote_servers>
    <cluster_name>
        <shard>
            <replica>
                <host>shard1-replica1-host</host>
                <port>9000</port>
            </replica>
            <replica>
                <host>shard1-replica2-host</host>
                <port>9000</port>
            </replica>
        </shard>
        <shard>
            <replica>
                <host>shard2-replica1-host</host>
                <port>9000</port>
            </replica>
            <replica>
                <host>shard2-replica2-host</host>
                <port>9000</port>
            </replica>
        </shard>
    </cluster_name>
</remote_servers>

<zookeeper>
    <node>
        <host>zookeeper1</host>
        <port>2181</port>
    </node>
    <node>
        <host>zookeeper2</host>
        <port>2181</port>
    </node>
</zookeeper>

Thus, each shard contains multiple replicas that are synchronized via ZooKeeper. The tables use the ReplicatedMergeTree engine with the path in ZooKeeper and the replica name, ensuring consistency and fault tolerance.