ClickHouse says its Postgres tools can link transactional databases with analytical queries, including a route that pushes 16 of 22 TPC-H queries to ClickHouse. The practical pitch is to keep Postgres for day-to-day transactions while sending analytical work to ClickHouse, with several tools smoothing the join between them.
ClickHouse Watch analysis
What happened
In a 7 October account of Postgres Summit US, ClickHouse describes four technologies for integrating Postgres and ClickHouse. The company says pgstatch records Postgres query metrics in ClickHouse for analysis, while pgclickhouse lets users query ClickHouse tables from Postgres using native Postgres SQL. ClickHouse says pgclickhouse currently pushes down 16 of 22 TPC-H queries.
The other tools target data movement and replication. ClickHouse says chdbhook enables imports and exports of data-lake files through Postgres’s COPY command, and WalShadow replicates data from Postgres to ClickHouse for what the company describes as sub-second visibility in analytical queries. These are the company’s descriptions; the post does not provide independent performance testing.
Why it matters
The appeal is less about replacing Postgres than giving it a companion for heavier analysis. Querying ClickHouse through familiar Postgres SQL could reduce friction for teams that already use both, while query metrics, file handling and replication address some of the practical plumbing between them.
The numbers give developers something concrete to investigate, but not a verdict on performance. The 16-of-22 figure is about TPC-H query pushdown, and the post does not supply comparative benchmark results for the full set of tools.
Our read
This is a useful integration story because it names the actual tools and what each is meant to do, rather than waving vaguely at a “unified data stack”. Teams already using Postgres and ClickHouse can now see which pieces might fit their workflow. Treat the performance claims as a starting point for testing against your own queries, not a ready-made benchmark win.
What to watch
- Whether pgclickhouse expands the number of TPC-H queries it can push down.
- How the tools perform on real workloads, beyond the examples in ClickHouse’s account.
- Whether ClickHouse publishes more detail on setup and compatibility across Postgres versions.
Discussion spark: If you already run Postgres and ClickHouse, which would make the biggest difference to your setup: query pushdown, easier data movement, or faster replication?
Sources and evidence
- Source update (7 October 2026, 16:56 UTC)
not affiliated with or endorsed by ClickHouse