ankitsultana opened a new issue, #14744: URL: https://github.com/apache/pinot/issues/14744
Was perf testing the and filter short-circuit PR #14700, and found out that the inverted index has the lowest priority and gets reordered so as to execute after all other filters. I think this might have happened after this PR #11137 which broke the bitmap based filter operator to two operators: bitmap-based and inverted-index. You can see a before-after the re-ordering in the image below. Also, the related code section is this: https://github.com/apache/pinot/blob/f8fa1a8a572f7668058bc29de23da586dc40c855/pinot-core/src/main/java/org/apache/pinot/core/operator/filter/FilterOperatorUtils.java#L222-L224 <img width="399" alt="Screenshot 2025-01-02 at 6 09 47 PM" src="https://github.com/user-attachments/assets/51c6e514-fcbc-4b66-adbb-38f6c0a64dc4"; /> <img width="364" alt="Screenshot 2025-01-02 at 6 10 03 PM" src="https://github.com/user-attachments/assets/adbc8195-f6ff-4505-ab79-5cb83be42c15"; /> @Jackie-Jiang : is this intentional or done in error? Also would you have a recommendation on the priority for the inverted index.. I think it should be at least equal to `LOW_PRIORITY`, and perhaps even `MEDIUM_PRIORITY`. -- 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: commits-unsubscr...@pinot.apache.org.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
