Use lower level Joda API in DateTimeTransformFunction

[richardstartin]

#8397 decreased allocation over 1000x and improved query time by ~30% for the query below evaluated over 2.5Bn events.

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:

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

[Jackie-Jiang]

Nice!

[codecov-commenter]

Codecov Report

Merging #8414 (14aaa99) into master (37fcfb8) will decrease coverage by 6.70%.
The diff coverage is 92.85%.

@@             Coverage Diff              @@
##             master    #8414      +/-   ##
============================================
- Coverage     70.77%   64.07%   -6.71%     
- Complexity     4278     4279       +1     
============================================
  Files          1655     1610      -45     
  Lines         86607    84784    -1823     
  Branches      13064    12864     -200     
============================================
- Hits          61296    54324    -6972     
- Misses        21068    26536    +5468     
+ Partials       4243     3924     -319     

Flag	Coverage Δ
integration1	?
integration2	?
unittests1	66.98% <92.85%> (-0.02%)	⬇️
unittests2	14.13% <0.00%> (+0.04%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
.../transform/function/DateTimeTransformFunction.java	89.24% <92.85%> (+0.63%)	⬆️
...va/org/apache/pinot/core/routing/RoutingTable.java	0.00% <0.00%> (-100.00%)	⬇️
...va/org/apache/pinot/common/config/NettyConfig.java	0.00% <0.00%> (-100.00%)	⬇️
...a/org/apache/pinot/common/metrics/MinionMeter.java	0.00% <0.00%> (-100.00%)	⬇️
...g/apache/pinot/common/metrics/ControllerMeter.java	0.00% <0.00%> (-100.00%)	⬇️
.../apache/pinot/common/metrics/BrokerQueryPhase.java	0.00% <0.00%> (-100.00%)	⬇️
.../apache/pinot/common/metrics/MinionQueryPhase.java	0.00% <0.00%> (-100.00%)	⬇️
...apache/pinot/common/helix/ExtraInstanceConfig.java	0.00% <0.00%> (-100.00%)	⬇️
...ache/pinot/server/access/AccessControlFactory.java	0.00% <0.00%> (-100.00%)	⬇️
...he/pinot/common/messages/SegmentReloadMessage.java	0.00% <0.00%> (-100.00%)	⬇️
... and 386 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 37fcfb8...14aaa99. Read the comment docs.

[xiangfu0]

Wow! Awesome!