songwdfu commented on PR #16308:
URL: https://github.com/apache/pinot/pull/16308#issuecomment-3105189486

   Some more benchmarking results on only the combine phase
   ```
   Benchmark                                                                    
   (_limit)  (_numRecordsPerSegment)  (_numSegments)  (_numThreads)             
                                                   (_query)  Mode  Cnt       
Score        Error  Units
   
BenchmarkPairwiseCombineOrderByGroupBy.concurrentIndexedTableForCombineGroupBy  
     500                    10000              20              4  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3  
119380.523 ± 115291.440  us/op
   
BenchmarkPairwiseCombineOrderByGroupBy.concurrentIndexedTableForCombineGroupBy  
     500                    10000              20              8  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3   
98326.437 ±  60942.907  us/op
   
BenchmarkPairwiseCombineOrderByGroupBy.concurrentIndexedTableForCombineGroupBy  
     500                    10000              20             16  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3  
111978.615 ±  57517.005  us/op
   
BenchmarkPairwiseCombineOrderByGroupBy.concurrentIndexedTableForCombineGroupBy  
    1000                    10000              20              4  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3  
112789.972 ± 123014.708  us/op
   
BenchmarkPairwiseCombineOrderByGroupBy.concurrentIndexedTableForCombineGroupBy  
    1000                    10000              20              8  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3  
115868.932 ±  30354.294  us/op
   
BenchmarkPairwiseCombineOrderByGroupBy.concurrentIndexedTableForCombineGroupBy  
    1000                    10000              20             16  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3  
105976.259 ± 188163.666  us/op
   BenchmarkPairwiseCombineOrderByGroupBy.sortedPairwiseCombineGroupBy          
        500                    10000              20              4  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3     
559.709 ±    820.137  us/op
   BenchmarkPairwiseCombineOrderByGroupBy.sortedPairwiseCombineGroupBy          
        500                    10000              20              8  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3     
516.331 ±    868.167  us/op
   BenchmarkPairwiseCombineOrderByGroupBy.sortedPairwiseCombineGroupBy          
        500                    10000              20             16  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3     
499.416 ±    642.365  us/op
   BenchmarkPairwiseCombineOrderByGroupBy.sortedPairwiseCombineGroupBy          
       1000                    10000              20              4  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3     
847.883 ±    348.402  us/op
   BenchmarkPairwiseCombineOrderByGroupBy.sortedPairwiseCombineGroupBy          
       1000                    10000              20              8  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3     
872.678 ±   2480.963  us/op
   BenchmarkPairwiseCombineOrderByGroupBy.sortedPairwiseCombineGroupBy          
       1000                    10000              20             16  SELECT 
sum(m1), max(m2) FROM testTable GROUP BY d1, d2 ORDER BY d1, d2  avgt    3     
761.380 ±     61.637  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: [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