[GitHub] [pinot] richardstartin opened a new pull request #8414: Use lower level Joda API

[GitBox]

richardstartin opened a new pull request #8414:
URL: https://github.com/apache/pinot/pull/8414


   #8397 decreased allocation over 1000x and improved query time by ~30% for 
the query below evaluated over 2.5Bn events.
   ```sql
   select count(*), year(event_time) as y, month(event_time) as m from 
githubEvents group by y, m
   ```
   
   The high level API leads to looking up lots of Chronology objects:
   
   <img width="1609" alt="Screenshot 2022-03-25 at 20 53 42" 
src="https://user-images.githubusercontent.com/16439049/160203730-c3b31e72-3e45-44e7-a1c2-cc1bd0af003c.png";>
   
   
   Joda has a slightly lower level API which allows 10-15% speedup by avoiding 
this, without taking on too much complexity.
   ```
   Benchmark               (_numRows)                                           
                                  (_query)  (_scenario)  Mode  Cnt      Score   
    Error  Units
   BenchmarkQueries.query     1500000  select count(*), year(INT_COL) as y, 
month(INT_COL) as m from MyTable group by y, m   EXP(0.001)  avgt    5  
59405.445 ± 2239.678  us/op
   BenchmarkQueries.query     1500000  select count(*), year(INT_COL) as y, 
month(INT_COL) as m from MyTable group by y, m     EXP(0.5)  avgt    5  
60158.445 ± 1162.191  us/op
   BenchmarkQueries.query     1500000  select count(*), year(INT_COL) as y, 
month(INT_COL) as m from MyTable group by y, m   EXP(0.999)  avgt    5  
59735.354 ± 1363.752  us/op
   ```
   
   ```
   Benchmark               (_numRows)                                           
                                  (_query)  (_scenario)  Mode  Cnt      Score   
   Error  Units
   BenchmarkQueries.query     1500000  select count(*), year(INT_COL) as y, 
month(INT_COL) as m from MyTable group by y, m   EXP(0.001)  avgt    5  
53623.538 ± 1757.387  us/op
   BenchmarkQueries.query     1500000  select count(*), year(INT_COL) as y, 
month(INT_COL) as m from MyTable group by y, m     EXP(0.5)  avgt    5  
54116.859 ± 2684.261  us/op
   BenchmarkQueries.query     1500000  select count(*), year(INT_COL) as y, 
month(INT_COL) as m from MyTable group by y, m   EXP(0.999)  avgt    5  
54359.302 ± 1790.208  us/op
   ```


-- 
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

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