chdb VS duckdb

Hey HN, creator of DB Pilot here.

I first announced DB Pilot on HN back in April: https://news.ycombinator.com/item?id=35761979.

Since then a lot has improved: More databases are supported, most of the product can now be used for free, and most importantly:

The app now comes with an analytics workspace powered by an embedded ClickHouse instance, running locally on your machine. This allows you to query local files, files on S3, PostgreSQL, SQLite & more - and all of those at once.

Embedding ClickHouse was possible thanks to chDB (https://github.com/chdb-io/chdb). A recent discussion on HN about it: https://news.ycombinator.com/item?id=37985005

I recommend using ClickHouse instead of DuckDB.

It has been around since 2016, and it covers and extends the feature set of DuckDB with a huge margin. Worth noting that it never has breaking changes in its table format MergeTree.

I'm tracking the progress of DuckDB and see that it is modeled after ClickHouse, but does not approach it in terms of feature completeness, stability, or performance.

The closest to DuckDB option is to use its self-contained version, clickhouse-local: https://clickhouse.com/blog/extracting-converting-querying-l... or an embedded version, chdb: https://github.com/chdb-io/chdb

Different beasts, but if by any chance you love ClickHouse already and just want to run OLAP queries in-process, there's chdb: https://github.com/chdb-io/chdb

> Can you embed it in Python as a library?

https://github.com/chdb-io/chdb

    pip install chdb

New kid on the block that I prefer over DuckDB is CHDB (https://github.com/chdb-io/chdb). Embedded ClickHouse, so once you out grow your laptop you can simply move to an actual OLAP that's Open-source.

I'm not sure if the fix is reassuring or not: https://github.com/duckdb/duckdb/pull/9411/files

It's an interesting question!

The problem is that the data is spread everywhere - no choice about that. So with that in mind, how do you query that data? Today, the idea is that you HAVE to put it into a central location. With tools like Bacalhau[1] and DuckDB [2], you no longer have to - a single query can be sharded amongst all your data - EFFECTIVELY giving you a lot of what you want from a data lake.

It's not a replacement, but if you can do a few of these items WITHOUT moving the data, you will be able to see really significant cost and time savings.

[1] https://github.com/bacalhau-project/bacalhau

[2] https://github.com/duckdb/duckdb

[4] Switch to Push-Based Execution Model by Mytherin · Pull Request #2393 · duckdb/duckdb (github.com)

it depends on your query obviously.

In general, I did very deep benchmarking of pg, clickhouse and duckdb, and I sure didn't make stupid mistakes like this: https://news.ycombinator.com/item?id=36990831

My dataset has 50B rows and 2tb of data, and I think columnar dbs are very overhiped and I chose pg because:

- pg performance is acceptable, maybe 2-3x times slower than clickhouse and duckdb on some queries if pg is configured correctly and run on compressed storage

- clickhouse and duckdb start falling apart very fast because they specialized on very narrow type of queries: https://github.com/ClickHouse/ClickHouse/issues/47520 https://github.com/ClickHouse/ClickHouse/issues/47521 https://github.com/duckdb/duckdb/discussions/6696

This action installs duckdb with the version provided in input.

Duckdb: https://github.com/duckdb/duckdb - seems pretty popular, been keeping an eye on this for close to a year now.

The Parquet-Go library is very complex, not yet success to use it. So I ask whether DuckDB can provide API https://github.com/duckdb/duckdb/issues/7776