Abhishek Batra
12/12/2025, 1:19 PMkapa.ai
12/12/2025, 1:19 PMReplicatedMergeTree, ReplicatedReplacingMergeTree, Distributed, Null), or - add ON CLUSTER ... to the CREATE TABLE statements. From the available information: - The v2 connector automatically chooses a MergeTree-family engine internally based on sync mode, and this behavior is hard‑coded rather than user‑configurable.[\[ClickHouse v2 discussion](https://github.com/airbytehq/airbyte/discussions/62035)\] - Users have requested: - the ability to select engine types (MergeTree, ReplacingMergeTree, Null), - partition configuration, and - more control over table creation, but these are discussed only as enhancement requests, not as shipped features.[\[ClickHouse v2 discussion](https://github.com/airbytehq/airbyte/discussions/62035)\] - Another user explicitly raised the need to “create replicated tables using the clickhouse connector”; this is logged as an open issue, with no documented solution or configuration knob yet.[\[Replicated tables issue](https://github.com/airbytehq/airbyte/issues/70245)\] - Earlier posts about sending data to ClickHouse clusters show that, at least for the legacy connector, the answer was that it “only works with single instance” and recommended custom code changes if you need ON CLUSTER / ReplicatedMergeTree behavior.[\[Cluster forum post](https://discuss.airbyte.io/t/1026)\] So, with the current, documented connector, you cannot configure replicated tables or ON CLUSTER directly via Airbyte’s UI or standard ClickHouse destination settings. Achieving that would require modifying the connector code (a custom build) to change the table‑creation SQL and temporary-table behavior. Because the docs and issues don’t describe an officially supported pattern for this, I can’t provide a supported configuration snippet or toggle for replicated/clustered tables.