GitHub user MisterRaindrop added a comment to the discussion: [Proposal] Iceberg subsystem for datalake_fdw — design proposal
I don’t think simply loading, reading or writing data is an issue for Cloudberry—we can optimize performance to match that of local PAx tables. When compared with Apache Doris/Starrocks, the difference may lie in the execution engine: we are based on the PostgreSQL engine, which is different from a pure columnar engine. It might be better to build a separate dedicated columnar analytical engine? I recall there was a previous discussion topic about pushing down queries to DuckDB; I wonder if this could be helpful later on. hi @yjhjstz Do you have any good suggestions on this part? GitHub link: https://github.com/apache/cloudberry/discussions/1683#discussioncomment-16646231 ---- This is an automatically sent email for [email protected]. To unsubscribe, please send an email to: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
