Question 1

How do Mongo documents map to ClickHouse columns?

Accepted Answer

Arrays map to native typed Array columns, including nested arrays, which is the good case. Objects, maps and tuples are not recreated as native Map or Tuple types. On ClickHouse 24.8 and later they land in a JSON column, and on older servers they fall back to a String holding serialized JSON. An array of objects therefore usually arrives as an array of JSON values. Worth checking your server version before assuming which you will get.

Question 2

Does my ClickHouse version matter here?

Accepted Answer

For this connector more than most, yes. The dividing line is 24.8, because native JSON is only available from that version. Below it, every object in your documents becomes a String containing JSON that you parse at query time, which works but is noticeably less pleasant on a document heavy source like Mongo. If you are running an older self hosted cluster and this pairing matters, upgrading is worth doing before the first sync rather than after.

Question 3

What kind of MongoDB deployment does this need?

Accepted Answer

One that supports change streams, so a replica set or sharded cluster rather than a standalone server. Polytomic opens a cluster level change stream filtered to your namespaces and operations, using resume tokens so a run continues where it stopped, which means the oplog has to be retained long enough to resume after a pause. The connecting user needs to list databases and collections, read the selected collections and open that stream.

Question 4

Why did a field not appear in my table?

Accepted Answer

Almost certainly because it was not in the sample. Schema inference reads the hundred most recent documents in a collection with recursion off, so a field appearing only in older documents, or rarely, may not make the initial schema. It can be added later, and anything discovered afterwards is added with an ALTER TABLE ADD COLUMN rather than needing a rebuild. Test against a collection you know is irregular before trusting the result.

Question 5

Why put document data in ClickHouse?

Accepted Answer

Because the shape suits it and the workload suits it. Mongo is awkward for aggregation across collections and competes with your application when you try, while ClickHouse is built for large scans and keeps arrays as real columns. ETL into a data warehouse is how the documents get there, and teams running high volume event or product data tend to end up with ClickHouse as that store rather than a general purpose warehouse.

Sync MongoDB to ClickHouse

Sync MongoDB to ClickHouse with Polytomic

Arrays stay native

Documents as JSON

Deletes arrive as tombstones

Or sync ClickHouse to MongoDB

Trusted by businesses worldwide

Step-by-step integration guide

Connect MongoDB in Polytomic

Connect ClickHouse in Polytomic

Create a sync between MongoDB and ClickHouse

Map fields and enable the sync

Frequently asked questions

The sync and ETL platform that does it all

Powerful data sync

Advanced analytics

Enterprise security

Easy setup

Ready to connect MongoDB with ClickHouse?