Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[SPARK-25476][SPARK-25510][TEST] Refactor AggregateBenchmark and add a new trait to better support Dataset and DataFrame API #22484

Closed
wants to merge 12 commits into from
Closed

[SPARK-25476][SPARK-25510][TEST] Refactor AggregateBenchmark and add a new trait to better support Dataset and DataFrame API #22484

Show file tree
Hide file tree
Changes from 7 commits
Commits
Show all changes
12 commits
Select commit Hold shift + click to select a range
649f296
Refactor AggregateBenchmark
wangyum Sep 20, 2018
e0c6d09
Merge remote-tracking branch 'upstream/master' into SPARK-25476
wangyum Sep 21, 2018
42230b6
Add RunBenchmarkWithCodegen
wangyum Sep 22, 2018
2d778a4
Rename RunBenchmarkWithCodegen to SqlBasedBenchmark
wangyum Sep 24, 2018
536d4e9
Add withSQLConf to SqlBasedBenchmark
wangyum Sep 24, 2018
f783099
true.toString -> "true" and false.toString -> "false"
wangyum Sep 25, 2018
d5fecdc
Fix indentation
wangyum Sep 25, 2018
23f1bd8
Merge remote-tracking branch 'upstream/master' into SPARK-25476
wangyum Sep 26, 2018
0fae54d
use SQLHelper
wangyum Sep 26, 2018
45add92
Update result (#12)
dongjoon-hyun Sep 27, 2018
3439992
Fix string variable
wangyum Sep 27, 2018
6c46ad5
Address comments
wangyum Oct 1, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
154 changes: 154 additions & 0 deletions sql/core/benchmarks/AggregateBenchmark-results.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,154 @@
================================================================================================
aggregate without grouping
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

agg w/o group: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
agg w/o group wholestage off 40468 / 44934 51.8 19.3 1.0X
agg w/o group wholestage on 932 / 949 2249.6 0.4 43.4X


================================================================================================
stat functions
================================================================================================

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@davies Do you know how to generate there benchmark:

spark/sql/core/src/test/scala/org/apache/spark/sql/execution/benchmark/AggregateBenchmark.scala

Lines 70 to 78 in 3c3eebc

Using ImperativeAggregate (as implemented in Spark 1.6):
Intel(R) Core(TM) i7-4558U CPU @ 2.80GHz
stddev: Avg Time(ms) Avg Rate(M/s) Relative Rate
-------------------------------------------------------------------------------
stddev w/o codegen 2019.04 10.39 1.00 X
stddev w codegen 2097.29 10.00 0.96 X
kurtosis w/o codegen 2108.99 9.94 0.96 X
kurtosis w codegen 2090.69 10.03 0.97 X

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

stddev: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
stddev wholestage off 6664 / 6775 15.7 63.6 1.0X
stddev wholestage on 907 / 923 115.6 8.6 7.3X

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

kurtosis: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
kurtosis wholestage off 31339 / 31357 3.3 298.9 1.0X
kurtosis wholestage on 982 / 997 106.8 9.4 31.9X


================================================================================================
aggregate with linear keys
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

Aggregate w keys: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
codegen = F 9667 / 9936 8.7 115.2 1.0X
codegen = T hashmap = F 4537 / 5353 18.5 54.1 2.1X
codegen = T hashmap = T 924 / 946 90.8 11.0 10.5X


================================================================================================
aggregate with randomized keys
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

Aggregate w keys: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
codegen = F 9298 / 9300 9.0 110.8 1.0X
codegen = T hashmap = F 5357 / 5405 15.7 63.9 1.7X
codegen = T hashmap = T 1705 / 1860 49.2 20.3 5.5X


================================================================================================
aggregate with string key
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

Aggregate w string key: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
codegen = F 4019 / 4083 5.2 191.7 1.0X
codegen = T hashmap = F 2743 / 2797 7.6 130.8 1.5X
codegen = T hashmap = T 1907 / 1955 11.0 91.0 2.1X


================================================================================================
aggregate with decimal key
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

Aggregate w decimal key: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
codegen = F 3229 / 3313 6.5 154.0 1.0X
codegen = T hashmap = F 2732 / 2826 7.7 130.3 1.2X
codegen = T hashmap = T 435 / 447 48.2 20.7 7.4X


================================================================================================
aggregate with multiple key types
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

Aggregate w multiple keys: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
codegen = F 6326 / 6800 3.3 301.6 1.0X
codegen = T hashmap = F 3915 / 4050 5.4 186.7 1.6X
codegen = T hashmap = T 2972 / 3054 7.1 141.7 2.1X


================================================================================================
max function bytecode size of wholestagecodegen
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

max function bytecode size: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
codegen = F 436 / 518 1.5 664.6 1.0X
codegen = T hugeMethodLimit = 10000 259 / 406 2.5 394.8 1.7X
codegen = T hugeMethodLimit = 1500 458 / 733 1.4 698.4 1.0X


================================================================================================
cube
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

cube: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
cube wholestage off 2957 / 2985 1.8 564.1 1.0X
cube wholestage on 1146 / 1675 4.6 218.5 2.6X


================================================================================================
hash and BytesToBytesMap
================================================================================================

Java HotSpot(TM) 64-Bit Server VM 1.8.0_151-b12 on Mac OS X 10.12.6
Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz

BytesToBytesMap: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
UnsafeRowhash 224 / 244 93.8 10.7 1.0X
murmur3 hash 114 / 134 184.2 5.4 2.0X
fast hash 61 / 66 342.8 2.9 3.7X
arrayEqual 150 / 229 139.5 7.2 1.5X
Java HashMap (Long) 127 / 305 164.7 6.1 1.8X
Java HashMap (two ints) 99 / 156 212.6 4.7 2.3X
Java HashMap (UnsafeRow) 1048 / 1472 20.0 50.0 0.2X
LongToUnsafeRowMap (opt=false) 517 / 771 40.6 24.7 0.4X
LongToUnsafeRowMap (opt=true) 90 / 107 234.1 4.3 2.5X
BytesToBytesMap (off Heap) 1417 / 1451 14.8 67.6 0.2X
BytesToBytesMap (on Heap) 934 / 1014 22.4 44.6 0.2X
Aggregate HashMap 38 / 42 558.6 1.8 6.0X


Loading