> That is a fine reason to have a first impression and a bad reason to make an architecture decision.

The name "ClickHouse" is an artifact of its origin. It was built at Yandex for Yandex.Metrica, a web analytics product, and the name is a contraction of "clickstream data warehouse." It describes the original workload and has nothing to do with what the engine does now.

If naming is going to be a factor, most of this category is not going to please you. StarRocks and Cloudberry are not obviously better. Doris is a person's name that came out of a Baidu project called Palo. DuckDB is named after a duck. QuestDB is generic. The one you have been running for eight years is named after a fast-moving antelope, and Hadoop was named after a toy elephant belonging to the creator's son.

Setting that aside, my recommendation from the previous message stands and does not depend on which name you prefer. The two candidates worth your time are StarRocks, if you want MPP performance with a MySQL-protocol endpoint and more conventional SQL than ClickHouse offers, and PostgreSQL with columnar storage, if you would rather run something your team already understands and accept a lower performance ceiling that your data size probably never reaches anyway.