[ClickHouse] DistributedEngine configuration

[karencfv]

The ClickHouse replicated SQL makes use of DistributedEngine tables. While we don't have any shards in place as of now, it is a preventive measure in case we decide to add any in the future.

Currently, we write and read directly to/from the DistributedEngine tables which in turn write to the ReplicatedMergeTree tables. This works, but we should research if this is the best way forward, or if we should write to the ReplicatedMergeTree tables, and query the DistributedEngine tables.

Related ClickHouse documentation: https://clickhouse.com/docs/en/engines/table-engines/special/distributed#distributed-writing-data

cc @bnaecker

[andrewjstone]

I'm pretty sure we've settled on writing directly to the DistributedEngine tables, and that's what we do in all our tests. This allows fault tolerance when we do have shards so that the available nodes can be chosen.