clintropolis commented on PR #18612: URL: https://github.com/apache/druid/pull/18612#issuecomment-3384501413
>You have an existing datasource A with all your data. You have some queries that is not performant on this datasource due to the low rollup ratio (i.e. this datasource has a requirement for fine granularity and/or large number of dimensions). You may want to use a Compaction task to read from datasource A and write to datasource A_optimizedForSomeGroupBy (exact copy with change in dimensionSpec and/or granularitySpec) to power this query. In this case, you wouldn't need to specific the metricsSpec as you are not changing the metricsSpec. drive by comment (still digesting the main idea of this PR), you should check out [projections](https://github.com/apache/druid/issues/17481) for this use case, which allow building aggregated views inside of segments the same datasource to automatically do more performant queries for any query which can match the projection. Projections can be applied to existing datasources via compaction since https://github.com/apache/druid/pull/17803. Projections are still experimental, and existing released versions have some bugs which will be fixed in the upcoming 35 release, where we plan to document them as part of that release since they should be a lot more stable. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
