yashmayya opened a new pull request, #17058:
URL: https://github.com/apache/pinot/pull/17058

   - Like https://github.com/apache/pinot/pull/16980/, but for `LONG` columns.
   - The primary goal here is to avoid the precision loss that comes with using 
double based computation for super large long values.
   - Unlike the `STRING` case though, here we also need to take into account 
backward compatibility - when `MIN` / `MAX` / `SUM` are rewritten, they should 
still be able to use existing star-tree indexes that would have pre-aggregated 
double values.
   - As a future enhancement, we'll also introduce type specific value 
aggregators so that the pre-aggregated star-tree index values can be the 
appropriate type instead of always being double. This enhancement will also 
need to be backward and forward compatible.
   - Note that we'll no longer be able to rely on the stored type hack for 
aggregation function column when the above enhancement will be made and we'll 
also need to refactor the star-tree index matching logic a bit. For instance, 
`MINLONG` should be able to use a star-tree index built with pre-aggregated 
long values or double values.


-- 
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]

Reply via email to