dongjoon-hyun commented on code in PR #54313:
URL: https://github.com/apache/spark/pull/54313#discussion_r2805635387
##########
sql/core/benchmarks/DataSourceReadBenchmark-jdk21-results.txt:
##########
@@ -2,437 +2,437 @@
SQL Single Numeric Column Scan
================================================================================================
-OpenJDK 64-Bit Server VM 21.0.8+9-LTS on Linux 6.11.0-1018-azure
+OpenJDK 64-Bit Server VM 21.0.10+7-LTS on Linux 6.14.0-1017-azure
AMD EPYC 7763 64-Core Processor
SQL Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 10326 10410
118 1.5 656.5 1.0X
-SQL Json 7489 7659
241 2.1 476.1 1.4X
-SQL Json with UnsafeRow 8554 8559
6 1.8 543.9 1.2X
-SQL Parquet Vectorized: DataPageV1 83 95
10 190.2 5.3 124.9X
-SQL Parquet Vectorized: DataPageV2 91 101
9 173.0 5.8 113.6X
-SQL Parquet MR: DataPageV1 1922 1931
13 8.2 122.2 5.4X
-SQL Parquet MR: DataPageV2 1740 1743
4 9.0 110.7 5.9X
-SQL ORC Vectorized 139 147
8 113.4 8.8 74.4X
-SQL ORC MR 1483 1496
19 10.6 94.3 7.0X
-
-OpenJDK 64-Bit Server VM 21.0.8+9-LTS on Linux 6.11.0-1018-azure
+SQL CSV 10371 10431
85 1.5 659.4 1.0X
+SQL Json 7575 7729
218 2.1 481.6 1.4X
+SQL Json with UnsafeRow 8696 8712
24 1.8 552.9 1.2X
+SQL Parquet Vectorized: DataPageV1 88 95
7 178.6 5.6 117.7X
+SQL Parquet Vectorized: DataPageV2 93 102
8 168.3 5.9 111.0X
+SQL Parquet MR: DataPageV1 1759 1761
3 8.9 111.9 5.9X
+SQL Parquet MR: DataPageV2 1636 1644
11 9.6 104.0 6.3X
+SQL ORC Vectorized 139 145
7 113.4 8.8 74.8X
+SQL ORC MR 1561 1575
19 10.1 99.3 6.6X
+
+OpenJDK 64-Bit Server VM 21.0.10+7-LTS on Linux 6.14.0-1017-azure
AMD EPYC 7763 64-Core Processor
Parquet Reader Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 93 96
1 168.9 5.9 1.0X
-ParquetReader Vectorized: DataPageV2 108 109
2 146.1 6.8 0.9X
-ParquetReader Vectorized -> Row: DataPageV1 75 76
1 209.1 4.8 1.2X
-ParquetReader Vectorized -> Row: DataPageV2 89 90
1 176.9 5.7 1.0X
+ParquetReader Vectorized: DataPageV1 91 92
1 172.6 5.8 1.0X
+ParquetReader Vectorized: DataPageV2 105 107
2 149.2 6.7 0.9X
+ParquetReader Vectorized -> Row: DataPageV1 72 74
2 217.6 4.6 1.3X
+ParquetReader Vectorized -> Row: DataPageV2 88 89
1 178.8 5.6 1.0X
-OpenJDK 64-Bit Server VM 21.0.8+9-LTS on Linux 6.11.0-1018-azure
+OpenJDK 64-Bit Server VM 21.0.10+7-LTS on Linux 6.14.0-1017-azure
AMD EPYC 7763 64-Core Processor
SQL Single TINYINT Column Scan: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 9609 9629
27 1.6 611.0 1.0X
-SQL Json 9153 9169
22 1.7 581.9 1.0X
-SQL Json with UnsafeRow 9804 9805
2 1.6 623.3 1.0X
-SQL Parquet Vectorized: DataPageV1 92 103
8 170.8 5.9 104.4X
-SQL Parquet Vectorized: DataPageV2 94 107
8 167.8 6.0 102.5X
-SQL Parquet MR: DataPageV1 1949 1971
31 8.1 123.9 4.9X
-SQL Parquet MR: DataPageV2 1889 1896
9 8.3 120.1 5.1X
-SQL ORC Vectorized 138 142
6 113.9 8.8 69.6X
-SQL ORC MR 1643 1667
34 9.6 104.5 5.8X
-
-OpenJDK 64-Bit Server VM 21.0.8+9-LTS on Linux 6.11.0-1018-azure
+SQL CSV 9487 9506
27 1.7 603.2 1.0X
+SQL Json 8686 8698
17 1.8 552.2 1.1X
+SQL Json with UnsafeRow 9338 9341
5 1.7 593.7 1.0X
+SQL Parquet Vectorized: DataPageV1 93 103
7 168.5 5.9 101.6X
+SQL Parquet Vectorized: DataPageV2 91 105
9 171.9 5.8 103.7X
+SQL Parquet MR: DataPageV1 1947 1957
14 8.1 123.8 4.9X
+SQL Parquet MR: DataPageV2 1883 1892
12 8.4 119.7 5.0X
+SQL ORC Vectorized 111 125
9 141.5 7.1 85.3X
+SQL ORC MR 1788 1796
11 8.8 113.7 5.3X
+
+OpenJDK 64-Bit Server VM 21.0.10+7-LTS on Linux 6.14.0-1017-azure
AMD EPYC 7763 64-Core Processor
Parquet Reader Single TINYINT Column Scan: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 86 87
1 182.6 5.5 1.0X
-ParquetReader Vectorized: DataPageV2 86 88
1 182.8 5.5 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 51 53
2 310.3 3.2 1.7X
-ParquetReader Vectorized -> Row: DataPageV2 51 52
1 311.4 3.2 1.7X
+ParquetReader Vectorized: DataPageV1 77 78
1 203.7 4.9 1.0X
+ParquetReader Vectorized: DataPageV2 78 78
1 202.4 4.9 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 49 50
1 322.6 3.1 1.6X
+ParquetReader Vectorized -> Row: DataPageV2 49 51
2 320.9 3.1 1.6X
-OpenJDK 64-Bit Server VM 21.0.8+9-LTS on Linux 6.11.0-1018-azure
+OpenJDK 64-Bit Server VM 21.0.10+7-LTS on Linux 6.14.0-1017-azure
AMD EPYC 7763 64-Core Processor
SQL Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 10101 10121
28 1.6 642.2 1.0X
-SQL Json 9160 9164
6 1.7 582.4 1.1X
-SQL Json with UnsafeRow 9855 9856
1 1.6 626.6 1.0X
-SQL Parquet Vectorized: DataPageV1 137 143
4 114.7 8.7 73.6X
-SQL Parquet Vectorized: DataPageV2 145 153
7 108.7 9.2 69.8X
-SQL Parquet MR: DataPageV1 2068 2086
26 7.6 131.5 4.9X
-SQL Parquet MR: DataPageV2 2040 2069
41 7.7 129.7 5.0X
-SQL ORC Vectorized 140 151
8 112.3 8.9 72.1X
-SQL ORC MR 1726 1727
2 9.1 109.7 5.9X
-
-OpenJDK 64-Bit Server VM 21.0.8+9-LTS on Linux 6.11.0-1018-azure
+SQL CSV 9786 9788
2 1.6 622.2 1.0X
+SQL Json 9669 9673
6 1.6 614.7 1.0X
+SQL Json with UnsafeRow 10439 10446
11 1.5 663.7 0.9X
+SQL Parquet Vectorized: DataPageV1 134 140
4 117.7 8.5 73.2X
+SQL Parquet Vectorized: DataPageV2 152 155
3 103.4 9.7 64.4X
+SQL Parquet MR: DataPageV1 2016 2052
50 7.8 128.2 4.9X
+SQL Parquet MR: DataPageV2 2110 2127
25 7.5 134.1 4.6X
+SQL ORC Vectorized 161 163
2 98.0 10.2 61.0X
+SQL ORC MR 1722 1738
23 9.1 109.5 5.7X
+
+OpenJDK 64-Bit Server VM 21.0.10+7-LTS on Linux 6.14.0-1017-azure
AMD EPYC 7763 64-Core Processor
Parquet Reader Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 166 168
2 94.6 10.6 1.0X
-ParquetReader Vectorized: DataPageV2 182 184
2 86.5 11.6 0.9X
-ParquetReader Vectorized -> Row: DataPageV1 158 162
2 99.4 10.1 1.1X
-ParquetReader Vectorized -> Row: DataPageV2 173 177
3 91.0 11.0 1.0X
+ParquetReader Vectorized: DataPageV1 167 171
4 94.0 10.6 1.0X
+ParquetReader Vectorized: DataPageV2 193 197
2 81.5 12.3 0.9X
+ParquetReader Vectorized -> Row: DataPageV1 171 179
8 92.1 10.9 1.0X
+ParquetReader Vectorized -> Row: DataPageV2 203 204
2 77.6 12.9 0.8X
Review Comment:
This becomes slower slightly.
--
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]