You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@spark.apache.org by do...@apache.org on 2022/08/04 16:30:51 UTC
[spark] branch master updated: [SPARK-39872][SQL] Change to use `BytePackerForLong#unpack8Values` with Array input api in `VectorizedDeltaBinaryPackedReader`
This is an automated email from the ASF dual-hosted git repository.
dongjoon pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/spark.git
The following commit(s) were added to refs/heads/master by this push:
new 0de98fd3936 [SPARK-39872][SQL] Change to use `BytePackerForLong#unpack8Values` with Array input api in `VectorizedDeltaBinaryPackedReader`
0de98fd3936 is described below
commit 0de98fd39364d2704607b031ddec623d48459202
Author: yangjie01 <ya...@baidu.com>
AuthorDate: Thu Aug 4 09:30:27 2022 -0700
[SPARK-39872][SQL] Change to use `BytePackerForLong#unpack8Values` with Array input api in `VectorizedDeltaBinaryPackedReader`
### What changes were proposed in this pull request?
This pr use `BytePackerForLong#unpack8Values` with `Array` input api instead of `BytePackerForLong#unpack8Values` with `ByteBuffer` input api in `VectorizedDeltaBinaryPackedReader` to improve the performance of scanning `INT` and `BIGINT` type data from Parquet DataPage V2.
### Why are the changes needed?
`BytePackerForLong#unpack8Values` with `ByteBuffer` input api has hotspot in using `HeapByteBuffer#get(int)` due to `HeapByteBuffer#get(int)` needs boundary check every time it is called, this affects the performance of scanning `INT` and `BIGINT` type data from Parquet DataPage V2.
The results of scenario `SQL Single INT/BIGINT Column Scan` as follows:
**Before**
```
OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
Intel(R) Xeon(R) Platinum 8171M CPU 2.60GHz
SQL Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
SQL Parquet Vectorized: DataPageV1 192 203 12 82.0 12.2 93.5X
SQL Parquet Vectorized: DataPageV2 363 373 10 43.4 23.1 49.5X
OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
Intel(R) Xeon(R) Platinum 8171M CPU 2.60GHz
SQL Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
SQL Parquet Vectorized: DataPageV1 252 262 11 62.3 16.1 92.8X
SQL Parquet Vectorized: DataPageV2 530 537 10 29.7 33.7 44.2X
```
**After**
```
OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
Intel(R) Xeon(R) Platinum 8272CL CPU 2.60GHz
SQL Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
SQL Parquet Vectorized: DataPageV1 173 201 70 91.1 11.0 91.2X
SQL Parquet Vectorized: DataPageV2 266 272 9 59.2 16.9 59.2X
OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
Intel(R) Xeon(R) Platinum 8272CL CPU 2.60GHz
SQL Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
SQL Parquet Vectorized: DataPageV1 235 243 9 66.9 15.0 88.3X
SQL Parquet Vectorized: DataPageV2 398 404 6 39.5 25.3 52.1X
```
After this pr, for the scanning scenario of type `INT` and `BIGINT` data, the performance gap between Parquet DataPage V2 and Parquet DataPage V1 is reduced.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
- Pass GitHub Actions
Closes #37293 from LuciferYang/ParquetV2-int-and-long.
Lead-authored-by: yangjie01 <ya...@baidu.com>
Co-authored-by: YangJie <ya...@baidu.com>
Signed-off-by: Dongjoon Hyun <do...@apache.org>
---
.../DataSourceReadBenchmark-jdk11-results.txt | 634 ++++++++++-----------
.../DataSourceReadBenchmark-jdk17-results.txt | 634 ++++++++++-----------
.../benchmarks/DataSourceReadBenchmark-results.txt | 634 ++++++++++-----------
.../parquet/VectorizedDeltaBinaryPackedReader.java | 3 +-
4 files changed, 953 insertions(+), 952 deletions(-)
diff --git a/sql/core/benchmarks/DataSourceReadBenchmark-jdk11-results.txt b/sql/core/benchmarks/DataSourceReadBenchmark-jdk11-results.txt
index a54fc6358a9..cf09ea90f65 100644
--- a/sql/core/benchmarks/DataSourceReadBenchmark-jdk11-results.txt
+++ b/sql/core/benchmarks/DataSourceReadBenchmark-jdk11-results.txt
@@ -2,430 +2,430 @@
SQL Single Numeric Column Scan
================================================================================================
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 14300 14340 57 1.1 909.1 1.0X
-SQL Json 10582 10589 10 1.5 672.8 1.4X
-SQL Parquet Vectorized: DataPageV1 117 147 48 134.3 7.4 122.1X
-SQL Parquet Vectorized: DataPageV2 94 110 21 167.4 6.0 152.2X
-SQL Parquet MR: DataPageV1 2215 2253 53 7.1 140.8 6.5X
-SQL Parquet MR: DataPageV2 1915 1939 33 8.2 121.8 7.5X
-SQL ORC Vectorized 201 242 37 78.1 12.8 71.0X
-SQL ORC MR 1853 1858 7 8.5 117.8 7.7X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 12550 12660 156 1.3 797.9 1.0X
+SQL Json 8037 8064 38 2.0 511.0 1.6X
+SQL Parquet Vectorized: DataPageV1 133 163 24 117.8 8.5 94.0X
+SQL Parquet Vectorized: DataPageV2 101 126 14 155.3 6.4 123.9X
+SQL Parquet MR: DataPageV1 2245 2250 7 7.0 142.7 5.6X
+SQL Parquet MR: DataPageV2 2035 2042 10 7.7 129.4 6.2X
+SQL ORC Vectorized 181 224 51 86.9 11.5 69.3X
+SQL ORC MR 1520 1546 37 10.3 96.6 8.3X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 156 168 14 100.7 9.9 1.0X
-ParquetReader Vectorized: DataPageV2 142 148 11 111.1 9.0 1.1X
-ParquetReader Vectorized -> Row: DataPageV1 66 68 4 239.4 4.2 2.4X
-ParquetReader Vectorized -> Row: DataPageV2 51 53 3 307.2 3.3 3.0X
+ParquetReader Vectorized: DataPageV1 127 150 25 123.4 8.1 1.0X
+ParquetReader Vectorized: DataPageV2 109 116 13 144.6 6.9 1.2X
+ParquetReader Vectorized -> Row: DataPageV1 56 58 3 280.8 3.6 2.3X
+ParquetReader Vectorized -> Row: DataPageV2 37 39 3 427.3 2.3 3.5X
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single TINYINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 17497 17581 120 0.9 1112.4 1.0X
-SQL Json 12087 12100 19 1.3 768.4 1.4X
-SQL Parquet Vectorized: DataPageV1 160 237 64 98.3 10.2 109.3X
-SQL Parquet Vectorized: DataPageV2 159 207 45 98.7 10.1 109.8X
-SQL Parquet MR: DataPageV1 2467 2484 24 6.4 156.8 7.1X
-SQL Parquet MR: DataPageV2 2279 2285 9 6.9 144.9 7.7X
-SQL ORC Vectorized 246 288 50 64.0 15.6 71.1X
-SQL ORC MR 1908 1922 21 8.2 121.3 9.2X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 15203 15226 33 1.0 966.6 1.0X
+SQL Json 9652 9660 11 1.6 613.6 1.6X
+SQL Parquet Vectorized: DataPageV1 141 178 36 111.3 9.0 107.6X
+SQL Parquet Vectorized: DataPageV2 138 167 39 114.0 8.8 110.2X
+SQL Parquet MR: DataPageV1 2380 2394 19 6.6 151.3 6.4X
+SQL Parquet MR: DataPageV2 2216 2226 15 7.1 140.9 6.9X
+SQL ORC Vectorized 180 208 30 87.4 11.4 84.5X
+SQL ORC MR 1649 1654 8 9.5 104.8 9.2X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single TINYINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 278 307 39 56.6 17.7 1.0X
-ParquetReader Vectorized: DataPageV2 229 247 30 68.7 14.6 1.2X
-ParquetReader Vectorized -> Row: DataPageV1 123 171 56 128.4 7.8 2.3X
-ParquetReader Vectorized -> Row: DataPageV2 121 147 29 129.9 7.7 2.3X
+ParquetReader Vectorized: DataPageV1 186 195 9 84.6 11.8 1.0X
+ParquetReader Vectorized: DataPageV2 186 192 7 84.5 11.8 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 95 99 9 166.1 6.0 2.0X
+ParquetReader Vectorized -> Row: DataPageV2 95 98 4 165.8 6.0 2.0X
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 18223 18240 23 0.9 1158.6 1.0X
-SQL Json 12557 12598 57 1.3 798.4 1.5X
-SQL Parquet Vectorized: DataPageV1 228 259 28 68.9 14.5 79.8X
-SQL Parquet Vectorized: DataPageV2 323 364 53 48.6 20.6 56.4X
-SQL Parquet MR: DataPageV1 2817 2885 97 5.6 179.1 6.5X
-SQL Parquet MR: DataPageV2 2486 2499 19 6.3 158.0 7.3X
-SQL ORC Vectorized 279 317 26 56.5 17.7 65.4X
-SQL ORC MR 2111 2140 41 7.4 134.2 8.6X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 16558 16567 13 0.9 1052.7 1.0X
+SQL Json 10190 10201 16 1.5 647.9 1.6X
+SQL Parquet Vectorized: DataPageV1 159 215 50 98.7 10.1 103.9X
+SQL Parquet Vectorized: DataPageV2 185 223 43 85.1 11.7 89.6X
+SQL Parquet MR: DataPageV1 2777 2786 13 5.7 176.5 6.0X
+SQL Parquet MR: DataPageV2 2410 2426 23 6.5 153.2 6.9X
+SQL ORC Vectorized 220 256 36 71.6 14.0 75.4X
+SQL ORC MR 1829 1844 20 8.6 116.3 9.1X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 339 358 18 46.4 21.5 1.0X
-ParquetReader Vectorized: DataPageV2 431 452 19 36.5 27.4 0.8X
-ParquetReader Vectorized -> Row: DataPageV1 334 348 13 47.0 21.3 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 441 446 6 35.7 28.0 0.8X
+ParquetReader Vectorized: DataPageV1 247 254 9 63.6 15.7 1.0X
+ParquetReader Vectorized: DataPageV2 275 282 9 57.2 17.5 0.9X
+ParquetReader Vectorized -> Row: DataPageV1 283 292 6 55.6 18.0 0.9X
+ParquetReader Vectorized -> Row: DataPageV2 261 293 30 60.3 16.6 0.9X
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 20053 20184 184 0.8 1275.0 1.0X
-SQL Json 13247 13257 15 1.2 842.2 1.5X
-SQL Parquet Vectorized: DataPageV1 239 261 14 65.8 15.2 83.9X
-SQL Parquet Vectorized: DataPageV2 480 516 36 32.8 30.5 41.8X
-SQL Parquet MR: DataPageV1 2859 2894 50 5.5 181.7 7.0X
-SQL Parquet MR: DataPageV2 2597 2597 1 6.1 165.1 7.7X
-SQL ORC Vectorized 373 389 13 42.2 23.7 53.8X
-SQL ORC MR 2185 2204 27 7.2 138.9 9.2X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 17997 18028 44 0.9 1144.2 1.0X
+SQL Json 10502 10513 16 1.5 667.7 1.7X
+SQL Parquet Vectorized: DataPageV1 153 188 36 102.8 9.7 117.6X
+SQL Parquet Vectorized: DataPageV2 324 395 40 48.6 20.6 55.6X
+SQL Parquet MR: DataPageV1 2688 2740 75 5.9 170.9 6.7X
+SQL Parquet MR: DataPageV2 2464 2483 26 6.4 156.7 7.3X
+SQL ORC Vectorized 288 348 54 54.7 18.3 62.6X
+SQL ORC MR 1795 1820 35 8.8 114.1 10.0X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 367 377 9 42.9 23.3 1.0X
-ParquetReader Vectorized: DataPageV2 606 611 6 26.0 38.5 0.6X
-ParquetReader Vectorized -> Row: DataPageV1 330 341 11 47.6 21.0 1.1X
-ParquetReader Vectorized -> Row: DataPageV2 566 575 13 27.8 36.0 0.6X
+ParquetReader Vectorized: DataPageV1 278 286 7 56.5 17.7 1.0X
+ParquetReader Vectorized: DataPageV2 453 461 9 34.7 28.8 0.6X
+ParquetReader Vectorized -> Row: DataPageV1 304 317 11 51.7 19.3 0.9X
+ParquetReader Vectorized -> Row: DataPageV2 474 482 9 33.2 30.1 0.6X
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 26123 26201 110 0.6 1660.8 1.0X
-SQL Json 16215 16220 7 1.0 1030.9 1.6X
-SQL Parquet Vectorized: DataPageV1 324 354 35 48.5 20.6 80.6X
-SQL Parquet Vectorized: DataPageV2 722 731 10 21.8 45.9 36.2X
-SQL Parquet MR: DataPageV1 2947 2960 19 5.3 187.4 8.9X
-SQL Parquet MR: DataPageV2 2745 2761 23 5.7 174.5 9.5X
-SQL ORC Vectorized 564 589 35 27.9 35.8 46.3X
-SQL ORC MR 2488 2490 3 6.3 158.2 10.5X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 22833 22857 34 0.7 1451.7 1.0X
+SQL Json 12611 12620 13 1.2 801.8 1.8X
+SQL Parquet Vectorized: DataPageV1 224 282 27 70.3 14.2 102.1X
+SQL Parquet Vectorized: DataPageV2 491 493 2 32.1 31.2 46.5X
+SQL Parquet MR: DataPageV1 2811 2817 9 5.6 178.7 8.1X
+SQL Parquet MR: DataPageV2 2642 2693 71 6.0 168.0 8.6X
+SQL ORC Vectorized 409 454 30 38.5 26.0 55.9X
+SQL ORC MR 2056 2064 11 7.7 130.7 11.1X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 407 422 27 38.7 25.9 1.0X
-ParquetReader Vectorized: DataPageV2 802 813 16 19.6 51.0 0.5X
-ParquetReader Vectorized -> Row: DataPageV1 446 456 13 35.3 28.3 0.9X
-ParquetReader Vectorized -> Row: DataPageV2 838 845 11 18.8 53.3 0.5X
+ParquetReader Vectorized: DataPageV1 350 366 25 45.0 22.2 1.0X
+ParquetReader Vectorized: DataPageV2 546 599 37 28.8 34.7 0.6X
+ParquetReader Vectorized -> Row: DataPageV1 372 385 12 42.3 23.6 0.9X
+ParquetReader Vectorized -> Row: DataPageV2 565 576 15 27.8 35.9 0.6X
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single FLOAT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 20982 21059 109 0.7 1334.0 1.0X
-SQL Json 15640 15688 68 1.0 994.4 1.3X
-SQL Parquet Vectorized: DataPageV1 204 241 16 77.1 13.0 102.8X
-SQL Parquet Vectorized: DataPageV2 197 241 27 79.8 12.5 106.4X
-SQL Parquet MR: DataPageV1 2816 2838 31 5.6 179.0 7.5X
-SQL Parquet MR: DataPageV2 2528 2560 46 6.2 160.7 8.3X
-SQL ORC Vectorized 549 554 5 28.6 34.9 38.2X
-SQL ORC MR 2351 2371 29 6.7 149.5 8.9X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 18495 18535 57 0.9 1175.9 1.0X
+SQL Json 12176 12190 20 1.3 774.1 1.5X
+SQL Parquet Vectorized: DataPageV1 155 208 37 101.7 9.8 119.6X
+SQL Parquet Vectorized: DataPageV2 152 201 38 103.3 9.7 121.5X
+SQL Parquet MR: DataPageV1 2639 2649 14 6.0 167.8 7.0X
+SQL Parquet MR: DataPageV2 2457 2489 46 6.4 156.2 7.5X
+SQL ORC Vectorized 450 456 6 34.9 28.6 41.1X
+SQL ORC MR 1931 1966 50 8.1 122.8 9.6X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single FLOAT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 331 344 20 47.5 21.0 1.0X
-ParquetReader Vectorized: DataPageV2 337 354 26 46.7 21.4 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 330 349 18 47.6 21.0 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 340 350 12 46.3 21.6 1.0X
+ParquetReader Vectorized: DataPageV1 280 284 11 56.2 17.8 1.0X
+ParquetReader Vectorized: DataPageV2 276 284 11 57.0 17.5 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 251 310 54 62.6 16.0 1.1X
+ParquetReader Vectorized -> Row: DataPageV2 252 300 33 62.5 16.0 1.1X
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single DOUBLE Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 26999 27012 19 0.6 1716.5 1.0X
-SQL Json 21266 21334 95 0.7 1352.1 1.3X
-SQL Parquet Vectorized: DataPageV1 333 352 17 47.3 21.2 81.1X
-SQL Parquet Vectorized: DataPageV2 331 363 27 47.5 21.1 81.5X
-SQL Parquet MR: DataPageV1 2942 2973 43 5.3 187.1 9.2X
-SQL Parquet MR: DataPageV2 2771 2801 42 5.7 176.2 9.7X
-SQL ORC Vectorized 690 720 46 22.8 43.9 39.1X
-SQL ORC MR 2538 2552 20 6.2 161.3 10.6X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 23272 23314 60 0.7 1479.6 1.0X
+SQL Json 16273 16277 6 1.0 1034.6 1.4X
+SQL Parquet Vectorized: DataPageV1 281 299 24 56.0 17.8 82.9X
+SQL Parquet Vectorized: DataPageV2 283 293 9 55.6 18.0 82.3X
+SQL Parquet MR: DataPageV1 2843 2846 4 5.5 180.7 8.2X
+SQL Parquet MR: DataPageV2 2692 2717 36 5.8 171.2 8.6X
+SQL ORC Vectorized 500 538 40 31.4 31.8 46.5X
+SQL ORC MR 2199 2217 26 7.2 139.8 10.6X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single DOUBLE Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 411 428 19 38.3 26.1 1.0X
-ParquetReader Vectorized: DataPageV2 417 424 10 37.8 26.5 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 440 455 16 35.7 28.0 0.9X
-ParquetReader Vectorized -> Row: DataPageV2 439 445 7 35.8 27.9 0.9X
+ParquetReader Vectorized: DataPageV1 318 334 25 49.5 20.2 1.0X
+ParquetReader Vectorized: DataPageV2 318 334 24 49.4 20.2 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 356 364 8 44.2 22.6 0.9X
+ParquetReader Vectorized -> Row: DataPageV2 348 364 12 45.1 22.2 0.9X
================================================================================================
SQL Single Numeric Column Scan in Struct
================================================================================================
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single TINYINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 2917 2936 26 5.4 185.5 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 2864 2882 25 5.5 182.1 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 313 361 37 50.2 19.9 9.3X
-SQL Parquet MR: DataPageV1 3113 3136 32 5.1 197.9 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3822 3900 110 4.1 243.0 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 341 388 34 46.2 21.7 8.6X
-SQL Parquet MR: DataPageV2 3009 3029 28 5.2 191.3 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3827 3829 2 4.1 243.3 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 199 255 55 79.1 12.6 14.7X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2493 2502 13 6.3 158.5 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2551 2559 10 6.2 162.2 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 245 281 37 64.1 15.6 10.2X
+SQL Parquet MR: DataPageV1 2901 2904 5 5.4 184.4 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3449 3493 62 4.6 219.3 0.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 180 213 35 87.2 11.5 13.8X
+SQL Parquet MR: DataPageV2 2793 2794 3 5.6 177.6 0.9X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3303 3361 82 4.8 210.0 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 179 194 20 88.1 11.4 14.0X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single SMALLINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3062 3076 20 5.1 194.7 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3010 3086 109 5.2 191.3 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 561 568 8 28.0 35.7 5.5X
-SQL Parquet MR: DataPageV1 3431 3541 156 4.6 218.1 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4272 4301 40 3.7 271.6 0.7X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 262 306 69 60.1 16.6 11.7X
-SQL Parquet MR: DataPageV2 3087 3165 111 5.1 196.3 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3946 3988 59 4.0 250.9 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 440 495 34 35.8 28.0 7.0X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2674 2674 1 5.9 170.0 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2596 2598 3 6.1 165.0 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 412 477 42 38.2 26.2 6.5X
+SQL Parquet MR: DataPageV1 3240 3244 6 4.9 206.0 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3822 3846 35 4.1 243.0 0.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 192 255 27 82.1 12.2 13.9X
+SQL Parquet MR: DataPageV2 2983 2984 2 5.3 189.7 0.9X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3558 3564 9 4.4 226.2 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 273 320 33 57.7 17.3 9.8X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single INT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3191 3246 79 4.9 202.9 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3170 3182 18 5.0 201.5 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 599 607 9 26.2 38.1 5.3X
-SQL Parquet MR: DataPageV1 3318 3325 10 4.7 211.0 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4202 4213 16 3.7 267.1 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 261 293 59 60.3 16.6 12.2X
-SQL Parquet MR: DataPageV2 3164 3198 49 5.0 201.1 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3957 3989 45 4.0 251.6 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 581 596 13 27.1 37.0 5.5X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2640 2663 33 6.0 167.9 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2627 2655 41 6.0 167.0 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 482 505 21 32.7 30.6 5.5X
+SQL Parquet MR: DataPageV1 3211 3245 47 4.9 204.2 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3810 3827 25 4.1 242.2 0.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 237 318 73 66.3 15.1 11.1X
+SQL Parquet MR: DataPageV2 3015 3024 12 5.2 191.7 0.9X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3614 3620 8 4.4 229.8 0.7X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 438 456 11 35.9 27.8 6.0X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BIGINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3336 3382 64 4.7 212.1 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3470 3474 6 4.5 220.6 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 782 791 8 20.1 49.7 4.3X
-SQL Parquet MR: DataPageV1 3420 3427 10 4.6 217.4 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4453 4510 80 3.5 283.1 0.7X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 372 474 60 42.3 23.6 9.0X
-SQL Parquet MR: DataPageV2 3561 3605 62 4.4 226.4 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 4239 4263 33 3.7 269.5 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 744 844 87 21.1 47.3 4.5X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2953 2961 11 5.3 187.8 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2886 2923 52 5.5 183.5 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 684 692 10 23.0 43.5 4.3X
+SQL Parquet MR: DataPageV1 3272 3283 15 4.8 208.0 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3897 3924 38 4.0 247.7 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 421 431 12 37.4 26.7 7.0X
+SQL Parquet MR: DataPageV2 3124 3178 77 5.0 198.6 0.9X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3726 3734 12 4.2 236.9 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 549 561 13 28.7 34.9 5.4X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single FLOAT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3337 3383 65 4.7 212.1 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3384 3405 30 4.6 215.1 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 762 801 38 20.6 48.4 4.4X
-SQL Parquet MR: DataPageV1 3208 3215 9 4.9 204.0 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4174 4176 3 3.8 265.4 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 260 403 72 60.6 16.5 12.8X
-SQL Parquet MR: DataPageV2 3140 3142 3 5.0 199.6 1.1X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3904 3934 43 4.0 248.2 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 209 260 31 75.4 13.3 16.0X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2852 2863 16 5.5 181.3 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2796 2828 46 5.6 177.8 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 611 619 7 25.7 38.9 4.7X
+SQL Parquet MR: DataPageV1 3127 3128 1 5.0 198.8 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3647 3692 64 4.3 231.9 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 178 229 40 88.6 11.3 16.1X
+SQL Parquet MR: DataPageV2 2912 2956 62 5.4 185.2 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3495 3549 76 4.5 222.2 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 178 235 42 88.3 11.3 16.0X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single DOUBLE Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3472 3503 44 4.5 220.7 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3475 3496 30 4.5 220.9 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 902 915 15 17.4 57.4 3.8X
-SQL Parquet MR: DataPageV1 3522 3533 16 4.5 223.9 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4453 4460 10 3.5 283.1 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 344 468 71 45.7 21.9 10.1X
-SQL Parquet MR: DataPageV2 3428 3534 150 4.6 217.9 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 4327 4376 70 3.6 275.1 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 341 384 51 46.1 21.7 10.2X
+SQL ORC MR 2996 3003 10 5.2 190.5 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2976 2982 9 5.3 189.2 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 711 714 4 22.1 45.2 4.2X
+SQL Parquet MR: DataPageV1 3360 3419 83 4.7 213.6 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3992 4036 62 3.9 253.8 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 305 347 58 51.5 19.4 9.8X
+SQL Parquet MR: DataPageV2 3246 3285 55 4.8 206.4 0.9X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3830 3863 46 4.1 243.5 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 322 341 29 48.9 20.5 9.3X
================================================================================================
SQL Nested Column Scan
================================================================================================
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Nested Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 29696 30578 448 0.0 28320.1 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 30799 31043 210 0.0 29372.6 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 11724 12024 157 0.1 11180.6 2.5X
-SQL Parquet MR: DataPageV1 20626 20868 178 0.1 19670.1 1.4X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 21362 21712 279 0.0 20372.0 1.4X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 12904 13043 107 0.1 12306.6 2.3X
-SQL Parquet MR: DataPageV2 22283 23110 492 0.0 21251.0 1.3X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 23156 24031 604 0.0 22083.5 1.3X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 10373 10527 102 0.1 9892.8 2.9X
+SQL ORC MR 22492 22654 79 0.0 21450.0 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 22525 22687 88 0.0 21481.8 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 10057 10279 234 0.1 9590.9 2.2X
+SQL Parquet MR: DataPageV1 17658 17799 97 0.1 16840.1 1.3X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 18063 18379 162 0.1 17226.5 1.2X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 11495 11715 131 0.1 10962.4 2.0X
+SQL Parquet MR: DataPageV2 16853 16976 151 0.1 16072.0 1.3X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 17553 17683 192 0.1 16740.0 1.3X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 8962 9017 35 0.1 8546.8 2.5X
================================================================================================
Int and String Scan
================================================================================================
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Int and String Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 18072 18100 39 0.6 1723.5 1.0X
-SQL Json 14735 14934 281 0.7 1405.2 1.2X
-SQL Parquet Vectorized: DataPageV1 2643 2649 9 4.0 252.1 6.8X
-SQL Parquet Vectorized: DataPageV2 3101 3107 8 3.4 295.7 5.8X
-SQL Parquet MR: DataPageV1 5685 5687 2 1.8 542.2 3.2X
-SQL Parquet MR: DataPageV2 5549 5561 16 1.9 529.2 3.3X
-SQL ORC Vectorized 2712 2712 1 3.9 258.6 6.7X
-SQL ORC MR 4809 4824 21 2.2 458.6 3.8X
+SQL CSV 15556 15628 101 0.7 1483.6 1.0X
+SQL Json 11804 11811 10 0.9 1125.7 1.3X
+SQL Parquet Vectorized: DataPageV1 2265 2283 25 4.6 216.0 6.9X
+SQL Parquet Vectorized: DataPageV2 2635 2729 133 4.0 251.3 5.9X
+SQL Parquet MR: DataPageV1 4777 4783 9 2.2 455.5 3.3X
+SQL Parquet MR: DataPageV2 4621 4631 14 2.3 440.7 3.4X
+SQL ORC Vectorized 2228 2239 16 4.7 212.5 7.0X
+SQL ORC MR 3864 3865 1 2.7 368.5 4.0X
================================================================================================
Repeated String Scan
================================================================================================
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Repeated String: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 10015 10227 300 1.0 955.1 1.0X
-SQL Json 9514 9520 8 1.1 907.4 1.1X
-SQL Parquet Vectorized: DataPageV1 779 806 24 13.5 74.3 12.9X
-SQL Parquet Vectorized: DataPageV2 761 787 40 13.8 72.6 13.2X
-SQL Parquet MR: DataPageV1 2345 2349 5 4.5 223.6 4.3X
-SQL Parquet MR: DataPageV2 2244 2276 45 4.7 214.0 4.5X
-SQL ORC Vectorized 601 650 50 17.4 57.3 16.7X
-SQL ORC MR 2399 2410 15 4.4 228.8 4.2X
+SQL CSV 8947 8962 21 1.2 853.3 1.0X
+SQL Json 7039 7053 20 1.5 671.3 1.3X
+SQL Parquet Vectorized: DataPageV1 598 623 21 17.5 57.1 15.0X
+SQL Parquet Vectorized: DataPageV2 592 620 22 17.7 56.4 15.1X
+SQL Parquet MR: DataPageV1 2164 2174 13 4.8 206.4 4.1X
+SQL Parquet MR: DataPageV2 2096 2102 8 5.0 199.9 4.3X
+SQL ORC Vectorized 519 526 5 20.2 49.5 17.2X
+SQL ORC MR 1929 1945 23 5.4 184.0 4.6X
================================================================================================
Partitioned Table Scan
================================================================================================
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Partitioned Table: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------------
-Data column - CSV 24890 24917 38 0.6 1582.5 1.0X
-Data column - Json 16405 16431 36 1.0 1043.0 1.5X
-Data column - Parquet Vectorized: DataPageV1 275 360 51 57.2 17.5 90.5X
-Data column - Parquet Vectorized: DataPageV2 752 775 29 20.9 47.8 33.1X
-Data column - Parquet MR: DataPageV1 3647 3679 45 4.3 231.9 6.8X
-Data column - Parquet MR: DataPageV2 3267 3278 15 4.8 207.7 7.6X
-Data column - ORC Vectorized 513 559 53 30.7 32.6 48.5X
-Data column - ORC MR 2821 2856 50 5.6 179.3 8.8X
-Partition column - CSV 8627 8690 90 1.8 548.5 2.9X
-Partition column - Json 13278 13305 39 1.2 844.2 1.9X
-Partition column - Parquet Vectorized: DataPageV1 68 94 34 232.2 4.3 367.5X
-Partition column - Parquet Vectorized: DataPageV2 67 83 22 235.2 4.3 372.2X
-Partition column - Parquet MR: DataPageV1 1757 1761 5 8.9 111.7 14.2X
-Partition column - Parquet MR: DataPageV2 1756 1769 18 9.0 111.6 14.2X
-Partition column - ORC Vectorized 69 92 24 227.6 4.4 360.2X
-Partition column - ORC MR 1787 1802 21 8.8 113.6 13.9X
-Both columns - CSV 26045 26118 104 0.6 1655.9 1.0X
-Both columns - Json 16817 16822 7 0.9 1069.2 1.5X
-Both columns - Parquet Vectorized: DataPageV1 426 448 21 36.9 27.1 58.4X
-Both columns - Parquet Vectorized: DataPageV2 904 911 8 17.4 57.4 27.5X
-Both columns - Parquet MR: DataPageV1 3603 3606 4 4.4 229.1 6.9X
-Both columns - Parquet MR: DataPageV2 3435 3481 64 4.6 218.4 7.2X
-Both columns - ORC Vectorized 598 636 47 26.3 38.0 41.6X
-Both columns - ORC MR 2982 3016 48 5.3 189.6 8.3X
+Data column - CSV 22125 22135 13 0.7 1406.7 1.0X
+Data column - Json 12724 12735 16 1.2 809.0 1.7X
+Data column - Parquet Vectorized: DataPageV1 230 300 42 68.3 14.6 96.1X
+Data column - Parquet Vectorized: DataPageV2 502 541 35 31.3 31.9 44.0X
+Data column - Parquet MR: DataPageV1 3254 3262 11 4.8 206.9 6.8X
+Data column - Parquet MR: DataPageV2 3165 3175 14 5.0 201.2 7.0X
+Data column - ORC Vectorized 424 450 25 37.1 26.9 52.2X
+Data column - ORC MR 2460 2471 15 6.4 156.4 9.0X
+Partition column - CSV 6768 6768 0 2.3 430.3 3.3X
+Partition column - Json 10555 10555 1 1.5 671.0 2.1X
+Partition column - Parquet Vectorized: DataPageV1 42 69 24 371.0 2.7 521.9X
+Partition column - Parquet Vectorized: DataPageV2 42 62 25 377.4 2.6 530.9X
+Partition column - Parquet MR: DataPageV1 1759 1773 21 8.9 111.8 12.6X
+Partition column - Parquet MR: DataPageV2 1756 1757 1 9.0 111.6 12.6X
+Partition column - ORC Vectorized 44 65 26 360.8 2.8 507.6X
+Partition column - ORC MR 1519 1525 9 10.4 96.6 14.6X
+Both columns - CSV 22212 22233 29 0.7 1412.2 1.0X
+Both columns - Json 13321 13349 40 1.2 846.9 1.7X
+Both columns - Parquet Vectorized: DataPageV1 321 345 23 49.0 20.4 68.9X
+Both columns - Parquet Vectorized: DataPageV2 548 565 17 28.7 34.8 40.4X
+Both columns - Parquet MR: DataPageV1 3313 3322 13 4.7 210.6 6.7X
+Both columns - Parquet MR: DataPageV2 3251 3260 13 4.8 206.7 6.8X
+Both columns - ORC Vectorized 450 473 23 34.9 28.6 49.2X
+Both columns - ORC MR 2535 2540 8 6.2 161.1 8.7X
================================================================================================
String with Nulls Scan
================================================================================================
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (0.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 12458 12494 52 0.8 1188.1 1.0X
-SQL Json 13083 13092 13 0.8 1247.7 1.0X
-SQL Parquet Vectorized: DataPageV1 1728 1760 46 6.1 164.8 7.2X
-SQL Parquet Vectorized: DataPageV2 2372 2417 63 4.4 226.2 5.3X
-SQL Parquet MR: DataPageV1 5207 5213 9 2.0 496.6 2.4X
-SQL Parquet MR: DataPageV2 5258 5267 12 2.0 501.4 2.4X
-ParquetReader Vectorized: DataPageV1 1234 1253 27 8.5 117.7 10.1X
-ParquetReader Vectorized: DataPageV2 1975 1994 27 5.3 188.3 6.3X
-SQL ORC Vectorized 1391 1457 93 7.5 132.7 9.0X
-SQL ORC MR 4032 4114 115 2.6 384.5 3.1X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 10557 10562 6 1.0 1006.8 1.0X
+SQL Json 10797 10811 20 1.0 1029.7 1.0X
+SQL Parquet Vectorized: DataPageV1 1304 1370 92 8.0 124.4 8.1X
+SQL Parquet Vectorized: DataPageV2 1927 1942 22 5.4 183.8 5.5X
+SQL Parquet MR: DataPageV1 4352 4357 8 2.4 415.0 2.4X
+SQL Parquet MR: DataPageV2 4531 4534 4 2.3 432.1 2.3X
+ParquetReader Vectorized: DataPageV1 989 1011 32 10.6 94.3 10.7X
+ParquetReader Vectorized: DataPageV2 1582 1584 3 6.6 150.9 6.7X
+SQL ORC Vectorized 1117 1162 64 9.4 106.6 9.4X
+SQL ORC MR 3248 3261 18 3.2 309.8 3.2X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (50.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 8879 8913 48 1.2 846.8 1.0X
-SQL Json 10052 10087 50 1.0 958.6 0.9X
-SQL Parquet Vectorized: DataPageV1 1402 1452 72 7.5 133.7 6.3X
-SQL Parquet Vectorized: DataPageV2 1908 1950 59 5.5 181.9 4.7X
-SQL Parquet MR: DataPageV1 3922 3979 80 2.7 374.1 2.3X
-SQL Parquet MR: DataPageV2 3986 4060 105 2.6 380.1 2.2X
-ParquetReader Vectorized: DataPageV1 1448 1451 5 7.2 138.1 6.1X
-ParquetReader Vectorized: DataPageV2 1830 1874 62 5.7 174.6 4.9X
-SQL ORC Vectorized 1609 1610 2 6.5 153.4 5.5X
-SQL ORC MR 3701 3756 77 2.8 353.0 2.4X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 8102 8109 9 1.3 772.7 1.0X
+SQL Json 8334 8343 13 1.3 794.8 1.0X
+SQL Parquet Vectorized: DataPageV1 1081 1138 80 9.7 103.1 7.5X
+SQL Parquet Vectorized: DataPageV2 1468 1481 18 7.1 140.0 5.5X
+SQL Parquet MR: DataPageV1 3380 3390 14 3.1 322.4 2.4X
+SQL Parquet MR: DataPageV2 3498 3502 6 3.0 333.6 2.3X
+ParquetReader Vectorized: DataPageV1 994 996 2 10.5 94.8 8.1X
+ParquetReader Vectorized: DataPageV2 1366 1369 4 7.7 130.3 5.9X
+SQL ORC Vectorized 1327 1342 22 7.9 126.6 6.1X
+SQL ORC MR 3080 3137 81 3.4 293.8 2.6X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (95.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 6443 6467 34 1.6 614.5 1.0X
-SQL Json 6207 6261 76 1.7 591.9 1.0X
-SQL Parquet Vectorized: DataPageV1 351 365 12 29.9 33.5 18.3X
-SQL Parquet Vectorized: DataPageV2 428 441 10 24.5 40.8 15.1X
-SQL Parquet MR: DataPageV1 2411 2419 11 4.3 229.9 2.7X
-SQL Parquet MR: DataPageV2 2290 2336 65 4.6 218.4 2.8X
-ParquetReader Vectorized: DataPageV1 372 379 4 28.2 35.5 17.3X
-ParquetReader Vectorized: DataPageV2 446 456 7 23.5 42.5 14.4X
-SQL ORC Vectorized 520 546 38 20.2 49.6 12.4X
-SQL ORC MR 1960 1964 6 5.3 186.9 3.3X
+SQL CSV 6090 6100 14 1.7 580.8 1.0X
+SQL Json 5001 5030 41 2.1 476.9 1.2X
+SQL Parquet Vectorized: DataPageV1 259 273 11 40.4 24.7 23.5X
+SQL Parquet Vectorized: DataPageV2 327 338 10 32.0 31.2 18.6X
+SQL Parquet MR: DataPageV1 2227 2235 12 4.7 212.4 2.7X
+SQL Parquet MR: DataPageV2 2123 2125 2 4.9 202.5 2.9X
+ParquetReader Vectorized: DataPageV1 263 267 3 39.8 25.1 23.1X
+ParquetReader Vectorized: DataPageV2 329 333 9 31.9 31.4 18.5X
+SQL ORC Vectorized 429 477 63 24.4 40.9 14.2X
+SQL ORC MR 1694 1702 11 6.2 161.6 3.6X
================================================================================================
Single Column Scan From Wide Columns
================================================================================================
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 10 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 2779 2785 8 0.4 2650.7 1.0X
-SQL Json 3752 3756 7 0.3 3577.9 0.7X
-SQL Parquet Vectorized: DataPageV1 53 74 25 19.9 50.2 52.8X
-SQL Parquet Vectorized: DataPageV2 81 98 22 12.9 77.3 34.3X
-SQL Parquet MR: DataPageV1 240 274 53 4.4 228.8 11.6X
-SQL Parquet MR: DataPageV2 227 273 64 4.6 216.8 12.2X
-SQL ORC Vectorized 65 85 27 16.0 62.3 42.5X
-SQL ORC MR 186 211 30 5.6 177.7 14.9X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 2623 2630 11 0.4 2501.3 1.0X
+SQL Json 2935 2943 11 0.4 2799.3 0.9X
+SQL Parquet Vectorized: DataPageV1 45 72 30 23.1 43.3 57.8X
+SQL Parquet Vectorized: DataPageV2 60 79 22 17.5 57.1 43.8X
+SQL Parquet MR: DataPageV1 220 257 41 4.8 209.9 11.9X
+SQL Parquet MR: DataPageV2 214 237 29 4.9 203.7 12.3X
+SQL ORC Vectorized 54 70 19 19.3 51.9 48.2X
+SQL ORC MR 160 185 29 6.5 152.8 16.4X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 50 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 6783 6791 12 0.2 6468.4 1.0X
-SQL Json 14463 14530 95 0.1 13792.6 0.5X
-SQL Parquet Vectorized: DataPageV1 77 115 36 13.6 73.3 88.3X
-SQL Parquet Vectorized: DataPageV2 103 130 27 10.2 98.3 65.8X
-SQL Parquet MR: DataPageV1 264 287 25 4.0 252.2 25.6X
-SQL Parquet MR: DataPageV2 249 279 36 4.2 237.9 27.2X
-SQL ORC Vectorized 88 111 26 11.9 84.3 76.7X
-SQL ORC MR 210 229 25 5.0 200.0 32.3X
-
-OpenJDK 64-Bit Server VM 11.0.14+9-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 6033 6043 14 0.2 5753.5 1.0X
+SQL Json 11341 11377 51 0.1 10816.0 0.5X
+SQL Parquet Vectorized: DataPageV1 63 103 28 16.7 59.8 96.3X
+SQL Parquet Vectorized: DataPageV2 77 108 33 13.7 73.1 78.7X
+SQL Parquet MR: DataPageV1 239 272 40 4.4 227.9 25.2X
+SQL Parquet MR: DataPageV2 234 266 33 4.5 223.5 25.7X
+SQL ORC Vectorized 71 95 27 14.7 67.9 84.7X
+SQL ORC MR 180 212 51 5.8 171.8 33.5X
+
+OpenJDK 64-Bit Server VM 11.0.16+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 100 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 11800 11807 10 0.1 11253.5 1.0X
-SQL Json 28084 28365 397 0.0 26782.9 0.4X
-SQL Parquet Vectorized: DataPageV1 127 153 21 8.3 120.7 93.2X
-SQL Parquet Vectorized: DataPageV2 145 171 25 7.2 138.1 81.5X
-SQL Parquet MR: DataPageV1 321 343 34 3.3 305.8 36.8X
-SQL Parquet MR: DataPageV2 293 333 40 3.6 279.7 40.2X
-SQL ORC Vectorized 113 130 21 9.3 107.4 104.8X
-SQL ORC MR 238 254 26 4.4 226.9 49.6X
+SQL CSV 10420 10435 22 0.1 9937.4 1.0X
+SQL Json 21702 21842 199 0.0 20696.2 0.5X
+SQL Parquet Vectorized: DataPageV1 98 130 24 10.7 93.4 106.4X
+SQL Parquet Vectorized: DataPageV2 104 130 27 10.1 99.5 99.9X
+SQL Parquet MR: DataPageV1 283 345 51 3.7 269.7 36.8X
+SQL Parquet MR: DataPageV2 269 294 31 3.9 256.4 38.8X
+SQL ORC Vectorized 89 104 19 11.7 85.2 116.7X
+SQL ORC MR 199 211 24 5.3 189.9 52.3X
diff --git a/sql/core/benchmarks/DataSourceReadBenchmark-jdk17-results.txt b/sql/core/benchmarks/DataSourceReadBenchmark-jdk17-results.txt
index 816c1d536d7..7b933f5c053 100644
--- a/sql/core/benchmarks/DataSourceReadBenchmark-jdk17-results.txt
+++ b/sql/core/benchmarks/DataSourceReadBenchmark-jdk17-results.txt
@@ -2,430 +2,430 @@
SQL Single Numeric Column Scan
================================================================================================
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 16481 16582 144 1.0 1047.8 1.0X
-SQL Json 9314 9441 180 1.7 592.2 1.8X
-SQL Parquet Vectorized: DataPageV1 118 131 11 133.4 7.5 139.8X
-SQL Parquet Vectorized: DataPageV2 88 98 8 178.3 5.6 186.9X
-SQL Parquet MR: DataPageV1 2354 2357 4 6.7 149.7 7.0X
-SQL Parquet MR: DataPageV2 2041 2051 15 7.7 129.7 8.1X
-SQL ORC Vectorized 180 197 11 87.3 11.5 91.5X
-SQL ORC MR 1912 1929 23 8.2 121.6 8.6X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 12691 12816 176 1.2 806.9 1.0X
+SQL Json 8145 8182 53 1.9 517.8 1.6X
+SQL Parquet Vectorized: DataPageV1 102 124 17 154.1 6.5 124.3X
+SQL Parquet Vectorized: DataPageV2 82 102 14 190.7 5.2 153.8X
+SQL Parquet MR: DataPageV1 1818 1821 4 8.7 115.6 7.0X
+SQL Parquet MR: DataPageV2 1752 1767 21 9.0 111.4 7.2X
+SQL ORC Vectorized 168 180 12 93.8 10.7 75.7X
+SQL ORC MR 1644 1655 15 9.6 104.5 7.7X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 79 80 2 200.0 5.0 1.0X
-ParquetReader Vectorized: DataPageV2 55 57 4 284.8 3.5 1.4X
-ParquetReader Vectorized -> Row: DataPageV1 58 63 4 270.0 3.7 1.4X
-ParquetReader Vectorized -> Row: DataPageV2 34 36 1 461.6 2.2 2.3X
+ParquetReader Vectorized: DataPageV1 59 61 3 267.8 3.7 1.0X
+ParquetReader Vectorized: DataPageV2 43 47 3 363.0 2.8 1.4X
+ParquetReader Vectorized -> Row: DataPageV1 40 42 2 389.6 2.6 1.5X
+ParquetReader Vectorized -> Row: DataPageV2 29 30 1 551.0 1.8 2.1X
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single TINYINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 19148 19160 17 0.8 1217.4 1.0X
-SQL Json 10723 10738 21 1.5 681.7 1.8X
-SQL Parquet Vectorized: DataPageV1 153 168 22 103.0 9.7 125.4X
-SQL Parquet Vectorized: DataPageV2 150 191 32 104.9 9.5 127.8X
-SQL Parquet MR: DataPageV1 2525 2548 32 6.2 160.5 7.6X
-SQL Parquet MR: DataPageV2 2332 2376 63 6.7 148.2 8.2X
-SQL ORC Vectorized 216 236 13 72.8 13.7 88.7X
-SQL ORC MR 1800 1841 57 8.7 114.4 10.6X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 14988 15174 262 1.0 952.9 1.0X
+SQL Json 9270 9278 12 1.7 589.4 1.6X
+SQL Parquet Vectorized: DataPageV1 207 225 14 75.9 13.2 72.3X
+SQL Parquet Vectorized: DataPageV2 207 231 18 76.1 13.1 72.5X
+SQL Parquet MR: DataPageV1 2010 2014 6 7.8 127.8 7.5X
+SQL Parquet MR: DataPageV2 1891 1901 15 8.3 120.2 7.9X
+SQL ORC Vectorized 164 181 9 96.0 10.4 91.5X
+SQL ORC MR 1455 1466 16 10.8 92.5 10.3X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single TINYINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 155 178 27 101.4 9.9 1.0X
-ParquetReader Vectorized: DataPageV2 154 160 7 102.2 9.8 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 108 115 10 145.4 6.9 1.4X
-ParquetReader Vectorized -> Row: DataPageV2 107 114 8 146.6 6.8 1.4X
+ParquetReader Vectorized: DataPageV1 202 204 2 78.0 12.8 1.0X
+ParquetReader Vectorized: DataPageV2 201 204 3 78.2 12.8 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 162 186 41 96.9 10.3 1.2X
+ParquetReader Vectorized -> Row: DataPageV2 161 164 3 97.8 10.2 1.3X
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 19660 19817 222 0.8 1250.0 1.0X
-SQL Json 11209 11292 117 1.4 712.6 1.8X
-SQL Parquet Vectorized: DataPageV1 172 191 16 91.6 10.9 114.4X
-SQL Parquet Vectorized: DataPageV2 237 263 23 66.4 15.1 83.0X
-SQL Parquet MR: DataPageV1 2726 2740 20 5.8 173.3 7.2X
-SQL Parquet MR: DataPageV2 2397 2423 37 6.6 152.4 8.2X
-SQL ORC Vectorized 243 271 25 64.8 15.4 81.0X
-SQL ORC MR 1990 2009 28 7.9 126.5 9.9X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 16299 16316 25 1.0 1036.2 1.0X
+SQL Json 9698 9713 22 1.6 616.6 1.7X
+SQL Parquet Vectorized: DataPageV1 151 157 5 104.3 9.6 108.1X
+SQL Parquet Vectorized: DataPageV2 172 182 7 91.7 10.9 95.0X
+SQL Parquet MR: DataPageV1 2109 2109 0 7.5 134.1 7.7X
+SQL Parquet MR: DataPageV2 2003 2007 5 7.9 127.4 8.1X
+SQL ORC Vectorized 205 216 9 76.8 13.0 79.6X
+SQL ORC MR 1725 1726 2 9.1 109.7 9.4X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 303 334 45 52.0 19.2 1.0X
-ParquetReader Vectorized: DataPageV2 363 367 4 43.3 23.1 0.8X
-ParquetReader Vectorized -> Row: DataPageV1 281 289 7 56.1 17.8 1.1X
-ParquetReader Vectorized -> Row: DataPageV2 340 362 20 46.3 21.6 0.9X
+ParquetReader Vectorized: DataPageV1 218 222 3 72.2 13.8 1.0X
+ParquetReader Vectorized: DataPageV2 242 245 3 65.0 15.4 0.9X
+ParquetReader Vectorized -> Row: DataPageV1 206 209 2 76.5 13.1 1.1X
+ParquetReader Vectorized -> Row: DataPageV2 229 231 3 68.7 14.6 1.0X
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 20834 21008 247 0.8 1324.6 1.0X
-SQL Json 11797 11933 191 1.3 750.1 1.8X
-SQL Parquet Vectorized: DataPageV1 187 223 23 84.2 11.9 111.5X
-SQL Parquet Vectorized: DataPageV2 383 412 23 41.1 24.3 54.4X
-SQL Parquet MR: DataPageV1 2820 2836 22 5.6 179.3 7.4X
-SQL Parquet MR: DataPageV2 2554 2595 58 6.2 162.4 8.2X
-SQL ORC Vectorized 329 344 11 47.7 20.9 63.2X
-SQL ORC MR 2177 2184 10 7.2 138.4 9.6X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 17548 17605 80 0.9 1115.7 1.0X
+SQL Json 10446 10479 47 1.5 664.1 1.7X
+SQL Parquet Vectorized: DataPageV1 140 150 9 112.2 8.9 125.2X
+SQL Parquet Vectorized: DataPageV2 224 246 15 70.2 14.3 78.3X
+SQL Parquet MR: DataPageV1 2136 2159 33 7.4 135.8 8.2X
+SQL Parquet MR: DataPageV2 1792 1793 1 8.8 114.0 9.8X
+SQL ORC Vectorized 233 251 10 67.5 14.8 75.3X
+SQL ORC MR 1688 1688 0 9.3 107.3 10.4X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 296 310 11 53.2 18.8 1.0X
-ParquetReader Vectorized: DataPageV2 442 466 18 35.6 28.1 0.7X
-ParquetReader Vectorized -> Row: DataPageV1 233 274 20 67.5 14.8 1.3X
-ParquetReader Vectorized -> Row: DataPageV2 443 450 9 35.5 28.2 0.7X
+ParquetReader Vectorized: DataPageV1 229 234 5 68.7 14.6 1.0X
+ParquetReader Vectorized: DataPageV2 329 332 3 47.7 20.9 0.7X
+ParquetReader Vectorized -> Row: DataPageV1 232 234 2 67.9 14.7 1.0X
+ParquetReader Vectorized -> Row: DataPageV2 335 337 4 47.0 21.3 0.7X
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 27229 27275 65 0.6 1731.2 1.0X
-SQL Json 14383 14434 73 1.1 914.4 1.9X
-SQL Parquet Vectorized: DataPageV1 225 283 40 69.9 14.3 121.0X
-SQL Parquet Vectorized: DataPageV2 496 522 27 31.7 31.5 54.9X
-SQL Parquet MR: DataPageV1 2969 2974 8 5.3 188.7 9.2X
-SQL Parquet MR: DataPageV2 2643 2688 64 6.0 168.0 10.3X
-SQL ORC Vectorized 417 471 42 37.7 26.5 65.3X
-SQL ORC MR 2234 2291 81 7.0 142.0 12.2X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 22478 22492 20 0.7 1429.1 1.0X
+SQL Json 13024 13071 66 1.2 828.0 1.7X
+SQL Parquet Vectorized: DataPageV1 201 214 11 78.3 12.8 111.8X
+SQL Parquet Vectorized: DataPageV2 320 336 13 49.1 20.4 70.2X
+SQL Parquet MR: DataPageV1 2426 2437 15 6.5 154.3 9.3X
+SQL Parquet MR: DataPageV2 2173 2221 67 7.2 138.2 10.3X
+SQL ORC Vectorized 308 337 26 51.1 19.6 73.0X
+SQL ORC MR 1844 1886 59 8.5 117.2 12.2X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 350 360 14 45.0 22.2 1.0X
-ParquetReader Vectorized: DataPageV2 578 600 21 27.2 36.8 0.6X
-ParquetReader Vectorized -> Row: DataPageV1 338 353 19 46.5 21.5 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 569 590 24 27.6 36.2 0.6X
+ParquetReader Vectorized: DataPageV1 274 278 4 57.4 17.4 1.0X
+ParquetReader Vectorized: DataPageV2 376 406 23 41.8 23.9 0.7X
+ParquetReader Vectorized -> Row: DataPageV1 295 324 39 53.4 18.7 0.9X
+ParquetReader Vectorized -> Row: DataPageV2 392 400 5 40.1 24.9 0.7X
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single FLOAT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 22581 22671 127 0.7 1435.6 1.0X
-SQL Json 14674 14711 52 1.1 933.0 1.5X
-SQL Parquet Vectorized: DataPageV1 180 227 35 87.5 11.4 125.6X
-SQL Parquet Vectorized: DataPageV2 207 228 19 76.1 13.1 109.2X
-SQL Parquet MR: DataPageV1 2678 2705 39 5.9 170.3 8.4X
-SQL Parquet MR: DataPageV2 2490 2507 24 6.3 158.3 9.1X
-SQL ORC Vectorized 489 502 15 32.2 31.1 46.2X
-SQL ORC MR 2139 2221 117 7.4 136.0 10.6X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 18349 18413 91 0.9 1166.6 1.0X
+SQL Json 12102 12137 50 1.3 769.4 1.5X
+SQL Parquet Vectorized: DataPageV1 148 188 48 106.5 9.4 124.3X
+SQL Parquet Vectorized: DataPageV2 137 155 14 114.8 8.7 133.9X
+SQL Parquet MR: DataPageV1 2236 2236 1 7.0 142.1 8.2X
+SQL Parquet MR: DataPageV2 2109 2164 76 7.5 134.1 8.7X
+SQL ORC Vectorized 464 481 13 33.9 29.5 39.5X
+SQL ORC MR 1966 1996 42 8.0 125.0 9.3X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single FLOAT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 267 280 18 59.0 16.9 1.0X
-ParquetReader Vectorized: DataPageV2 297 306 11 53.0 18.9 0.9X
-ParquetReader Vectorized -> Row: DataPageV1 298 319 21 52.7 19.0 0.9X
-ParquetReader Vectorized -> Row: DataPageV2 299 305 5 52.6 19.0 0.9X
+ParquetReader Vectorized: DataPageV1 234 253 16 67.1 14.9 1.0X
+ParquetReader Vectorized: DataPageV2 225 267 30 69.9 14.3 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 251 259 5 62.6 16.0 0.9X
+ParquetReader Vectorized -> Row: DataPageV2 254 272 10 61.9 16.2 0.9X
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single DOUBLE Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 28301 28738 617 0.6 1799.4 1.0X
-SQL Json 19561 19756 277 0.8 1243.6 1.4X
-SQL Parquet Vectorized: DataPageV1 231 275 23 68.0 14.7 122.3X
-SQL Parquet Vectorized: DataPageV2 272 302 23 57.9 17.3 104.2X
-SQL Parquet MR: DataPageV1 3018 3075 81 5.2 191.9 9.4X
-SQL Parquet MR: DataPageV2 2795 2828 46 5.6 177.7 10.1X
-SQL ORC Vectorized 526 619 80 29.9 33.5 53.8X
-SQL ORC MR 2342 2379 52 6.7 148.9 12.1X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 23077 23260 260 0.7 1467.2 1.0X
+SQL Json 16423 16445 31 1.0 1044.1 1.4X
+SQL Parquet Vectorized: DataPageV1 205 241 30 76.6 13.1 112.4X
+SQL Parquet Vectorized: DataPageV2 214 255 26 73.5 13.6 107.9X
+SQL Parquet MR: DataPageV1 2407 2429 32 6.5 153.0 9.6X
+SQL Parquet MR: DataPageV2 2319 2333 20 6.8 147.4 10.0X
+SQL ORC Vectorized 559 581 20 28.2 35.5 41.3X
+SQL ORC MR 2119 2133 20 7.4 134.7 10.9X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single DOUBLE Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 369 387 16 42.6 23.4 1.0X
-ParquetReader Vectorized: DataPageV2 367 377 9 42.8 23.4 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 371 402 33 42.4 23.6 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 371 395 26 42.4 23.6 1.0X
+ParquetReader Vectorized: DataPageV1 318 329 10 49.5 20.2 1.0X
+ParquetReader Vectorized: DataPageV2 301 326 21 52.3 19.1 1.1X
+ParquetReader Vectorized -> Row: DataPageV1 321 328 4 49.1 20.4 1.0X
+ParquetReader Vectorized -> Row: DataPageV2 323 340 19 48.6 20.6 1.0X
================================================================================================
SQL Single Numeric Column Scan in Struct
================================================================================================
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single TINYINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 2932 2980 68 5.4 186.4 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3029 3046 24 5.2 192.6 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 382 431 74 41.1 24.3 7.7X
-SQL Parquet MR: DataPageV1 3352 3362 15 4.7 213.1 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4184 4203 26 3.8 266.0 0.7X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 226 252 12 69.7 14.3 13.0X
-SQL Parquet MR: DataPageV2 3261 3275 20 4.8 207.4 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3973 4067 133 4.0 252.6 0.7X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 231 258 20 68.0 14.7 12.7X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2377 2439 88 6.6 151.1 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2343 2370 39 6.7 148.9 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 281 313 24 56.0 17.9 8.5X
+SQL Parquet MR: DataPageV1 2583 2617 48 6.1 164.2 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3008 3031 32 5.2 191.2 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 233 243 9 67.4 14.8 10.2X
+SQL Parquet MR: DataPageV2 2493 2506 18 6.3 158.5 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2914 2936 31 5.4 185.2 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 233 257 26 67.6 14.8 10.2X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single SMALLINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3020 3046 37 5.2 192.0 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3019 3035 22 5.2 192.0 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 481 502 18 32.7 30.6 6.3X
-SQL Parquet MR: DataPageV1 3448 3451 4 4.6 219.2 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4298 4357 82 3.7 273.3 0.7X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 235 254 20 66.9 14.9 12.9X
-SQL Parquet MR: DataPageV2 3251 3262 15 4.8 206.7 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 4068 4080 17 3.9 258.6 0.7X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 432 450 16 36.4 27.5 7.0X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2372 2404 45 6.6 150.8 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2323 2389 94 6.8 147.7 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 387 396 8 40.6 24.6 6.1X
+SQL Parquet MR: DataPageV1 2717 2725 11 5.8 172.8 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3272 3284 18 4.8 208.0 0.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 162 200 27 96.9 10.3 14.6X
+SQL Parquet MR: DataPageV2 2524 2546 31 6.2 160.5 0.9X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3151 3179 41 5.0 200.3 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 358 387 25 44.0 22.7 6.6X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single INT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3034 3177 203 5.2 192.9 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3262 3279 23 4.8 207.4 0.9X
-SQL ORC Vectorized (Nested Column Enabled) 502 531 26 31.3 31.9 6.0X
-SQL Parquet MR: DataPageV1 3473 3499 37 4.5 220.8 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4282 4303 29 3.7 272.3 0.7X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 187 236 39 84.1 11.9 16.2X
-SQL Parquet MR: DataPageV2 3340 3344 6 4.7 212.4 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 4071 4103 46 3.9 258.8 0.7X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 505 515 15 31.2 32.1 6.0X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2178 2243 91 7.2 138.5 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2382 2436 76 6.6 151.4 0.9X
+SQL ORC Vectorized (Nested Column Enabled) 438 467 36 35.9 27.8 5.0X
+SQL Parquet MR: DataPageV1 2649 2673 34 5.9 168.4 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3263 3270 10 4.8 207.4 0.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 197 229 45 79.9 12.5 11.1X
+SQL Parquet MR: DataPageV2 2645 2652 10 5.9 168.2 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3245 3248 4 4.8 206.3 0.7X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 401 417 21 39.2 25.5 5.4X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BIGINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3273 3301 39 4.8 208.1 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3342 3353 15 4.7 212.5 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 629 651 27 25.0 40.0 5.2X
-SQL Parquet MR: DataPageV1 3649 3694 64 4.3 232.0 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4559 4615 79 3.4 289.9 0.7X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 313 328 11 50.2 19.9 10.4X
-SQL Parquet MR: DataPageV2 3441 3482 58 4.6 218.8 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 4141 4211 100 3.8 263.2 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 493 511 12 31.9 31.4 6.6X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2686 2688 4 5.9 170.8 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2652 2661 12 5.9 168.6 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 542 563 22 29.0 34.5 5.0X
+SQL Parquet MR: DataPageV1 2869 2873 5 5.5 182.4 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3559 3641 116 4.4 226.3 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 265 278 12 59.3 16.9 10.1X
+SQL Parquet MR: DataPageV2 2722 2726 5 5.8 173.1 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3290 3307 24 4.8 209.2 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 370 395 19 42.5 23.5 7.3X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single FLOAT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3345 3360 21 4.7 212.7 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3171 3183 16 5.0 201.6 1.1X
-SQL ORC Vectorized (Nested Column Enabled) 646 668 20 24.3 41.1 5.2X
-SQL Parquet MR: DataPageV1 3376 3388 17 4.7 214.6 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4268 4309 58 3.7 271.3 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 171 180 6 92.0 10.9 19.6X
-SQL Parquet MR: DataPageV2 3209 3211 2 4.9 204.0 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3964 3983 26 4.0 252.0 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 171 207 26 91.9 10.9 19.6X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2598 2682 119 6.1 165.2 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2581 2621 57 6.1 164.1 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 660 676 26 23.8 41.9 3.9X
+SQL Parquet MR: DataPageV1 2305 2351 65 6.8 146.6 1.1X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3167 3223 78 5.0 201.4 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 162 199 23 97.1 10.3 16.0X
+SQL Parquet MR: DataPageV2 2201 2390 267 7.1 140.0 1.2X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3083 3133 69 5.1 196.0 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 152 187 32 103.6 9.7 17.1X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single DOUBLE Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 3288 3350 88 4.8 209.0 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 3303 3319 23 4.8 210.0 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 804 869 57 19.6 51.1 4.1X
-SQL Parquet MR: DataPageV1 3697 3734 52 4.3 235.1 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 4447 4450 4 3.5 282.8 0.7X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 295 314 22 53.3 18.8 11.1X
-SQL Parquet MR: DataPageV2 3554 3557 5 4.4 225.9 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 4282 4317 49 3.7 272.3 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 285 338 44 55.3 18.1 11.6X
+SQL ORC MR 2714 2754 57 5.8 172.5 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2603 2679 108 6.0 165.5 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 666 675 13 23.6 42.3 4.1X
+SQL Parquet MR: DataPageV1 2920 2943 32 5.4 185.7 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3493 3496 4 4.5 222.1 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 255 280 20 61.6 16.2 10.6X
+SQL Parquet MR: DataPageV2 2824 2834 14 5.6 179.5 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3260 3278 24 4.8 207.3 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 246 281 32 63.9 15.6 11.0X
================================================================================================
SQL Nested Column Scan
================================================================================================
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Nested Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 22294 22576 168 0.0 21261.4 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 22303 22632 246 0.0 21269.9 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 10280 10354 66 0.1 9803.3 2.2X
-SQL Parquet MR: DataPageV1 15143 15423 220 0.1 14441.1 1.5X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 15730 15988 226 0.1 15001.7 1.4X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 10560 10846 150 0.1 10070.5 2.1X
-SQL Parquet MR: DataPageV2 15428 16012 432 0.1 14713.7 1.4X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 15739 15895 110 0.1 15009.6 1.4X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 8049 8414 309 0.1 7676.3 2.8X
+SQL ORC MR 18905 19054 88 0.1 18029.3 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 18830 18980 83 0.1 17958.1 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 9226 9284 30 0.1 8798.7 2.0X
+SQL Parquet MR: DataPageV1 11687 12034 280 0.1 11146.0 1.6X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 12766 13054 169 0.1 12174.6 1.5X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 9002 9376 267 0.1 8585.1 2.1X
+SQL Parquet MR: DataPageV2 12390 12512 70 0.1 11816.2 1.5X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 12568 12677 78 0.1 11985.5 1.5X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 6551 6735 157 0.2 6247.3 2.9X
================================================================================================
Int and String Scan
================================================================================================
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Int and String Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 18849 18860 16 0.6 1797.6 1.0X
-SQL Json 13433 13494 86 0.8 1281.1 1.4X
-SQL Parquet Vectorized: DataPageV1 2539 2605 93 4.1 242.2 7.4X
-SQL Parquet Vectorized: DataPageV2 2825 2840 22 3.7 269.4 6.7X
-SQL Parquet MR: DataPageV1 5447 5460 18 1.9 519.5 3.5X
-SQL Parquet MR: DataPageV2 5513 5554 59 1.9 525.7 3.4X
-SQL ORC Vectorized 2567 2567 1 4.1 244.8 7.3X
-SQL ORC MR 4588 4687 139 2.3 437.6 4.1X
+SQL CSV 15161 15162 0 0.7 1445.9 1.0X
+SQL Json 11120 11122 2 0.9 1060.5 1.4X
+SQL Parquet Vectorized: DataPageV1 1914 1914 1 5.5 182.5 7.9X
+SQL Parquet Vectorized: DataPageV2 2202 2216 20 4.8 210.0 6.9X
+SQL Parquet MR: DataPageV1 4292 4334 59 2.4 409.3 3.5X
+SQL Parquet MR: DataPageV2 4210 4237 38 2.5 401.5 3.6X
+SQL ORC Vectorized 2000 2024 35 5.2 190.7 7.6X
+SQL ORC MR 3635 3666 44 2.9 346.7 4.2X
================================================================================================
Repeated String Scan
================================================================================================
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Repeated String: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 11328 11341 18 0.9 1080.3 1.0X
-SQL Json 8491 8504 19 1.2 809.7 1.3X
-SQL Parquet Vectorized: DataPageV1 729 741 11 14.4 69.6 15.5X
-SQL Parquet Vectorized: DataPageV2 692 716 25 15.2 66.0 16.4X
-SQL Parquet MR: DataPageV1 2199 2211 17 4.8 209.7 5.2X
-SQL Parquet MR: DataPageV2 2067 2074 11 5.1 197.1 5.5X
-SQL ORC Vectorized 559 593 30 18.8 53.3 20.3X
-SQL ORC MR 2324 2365 58 4.5 221.6 4.9X
+SQL CSV 8913 9170 364 1.2 850.0 1.0X
+SQL Json 6173 6319 207 1.7 588.7 1.4X
+SQL Parquet Vectorized: DataPageV1 574 611 31 18.3 54.8 15.5X
+SQL Parquet Vectorized: DataPageV2 621 646 42 16.9 59.3 14.3X
+SQL Parquet MR: DataPageV1 1727 1846 168 6.1 164.7 5.2X
+SQL Parquet MR: DataPageV2 1579 1581 3 6.6 150.6 5.6X
+SQL ORC Vectorized 458 473 20 22.9 43.6 19.5X
+SQL ORC MR 1719 1795 107 6.1 164.0 5.2X
================================================================================================
Partitioned Table Scan
================================================================================================
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Partitioned Table: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------------
-Data column - CSV 27302 27476 246 0.6 1735.8 1.0X
-Data column - Json 15448 15503 79 1.0 982.1 1.8X
-Data column - Parquet Vectorized: DataPageV1 234 239 6 67.4 14.8 116.9X
-Data column - Parquet Vectorized: DataPageV2 548 580 27 28.7 34.8 49.8X
-Data column - Parquet MR: DataPageV1 3444 3557 160 4.6 219.0 7.9X
-Data column - Parquet MR: DataPageV2 3254 3255 1 4.8 206.9 8.4X
-Data column - ORC Vectorized 462 545 103 34.0 29.4 59.1X
-Data column - ORC MR 2910 2960 70 5.4 185.0 9.4X
-Partition column - CSV 7922 8000 111 2.0 503.6 3.4X
-Partition column - Json 11991 12013 32 1.3 762.4 2.3X
-Partition column - Parquet Vectorized: DataPageV1 62 70 8 252.6 4.0 438.4X
-Partition column - Parquet Vectorized: DataPageV2 59 65 7 265.7 3.8 461.3X
-Partition column - Parquet MR: DataPageV1 1737 1831 133 9.1 110.4 15.7X
-Partition column - Parquet MR: DataPageV2 1753 1781 39 9.0 111.5 15.6X
-Partition column - ORC Vectorized 62 70 10 253.5 3.9 440.0X
-Partition column - ORC MR 1723 1731 11 9.1 109.6 15.8X
-Both columns - CSV 27325 27376 72 0.6 1737.3 1.0X
-Both columns - Json 15499 15606 151 1.0 985.4 1.8X
-Both columns - Parquet Vectorized: DataPageV1 369 394 23 42.7 23.4 74.1X
-Both columns - Parquet Vectorized: DataPageV2 583 622 30 27.0 37.1 46.8X
-Both columns - Parquet MR: DataPageV1 3641 3699 82 4.3 231.5 7.5X
-Both columns - Parquet MR: DataPageV2 3398 3456 83 4.6 216.0 8.0X
-Both columns - ORC Vectorized 533 564 37 29.5 33.9 51.2X
-Both columns - ORC MR 2918 2930 16 5.4 185.5 9.4X
+Data column - CSV 21176 21402 319 0.7 1346.3 1.0X
+Data column - Json 12134 12262 181 1.3 771.5 1.7X
+Data column - Parquet Vectorized: DataPageV1 194 218 17 81.2 12.3 109.4X
+Data column - Parquet Vectorized: DataPageV2 365 417 35 43.1 23.2 58.0X
+Data column - Parquet MR: DataPageV1 2665 2807 202 5.9 169.4 7.9X
+Data column - Parquet MR: DataPageV2 2523 2612 125 6.2 160.4 8.4X
+Data column - ORC Vectorized 411 440 16 38.3 26.1 51.5X
+Data column - ORC MR 2278 2281 4 6.9 144.9 9.3X
+Partition column - CSV 5972 6160 266 2.6 379.7 3.5X
+Partition column - Json 9695 9843 210 1.6 616.4 2.2X
+Partition column - Parquet Vectorized: DataPageV1 37 43 7 425.5 2.4 572.8X
+Partition column - Parquet Vectorized: DataPageV2 37 43 6 421.4 2.4 567.4X
+Partition column - Parquet MR: DataPageV1 1296 1301 6 12.1 82.4 16.3X
+Partition column - Parquet MR: DataPageV2 1286 1291 8 12.2 81.7 16.5X
+Partition column - ORC Vectorized 38 42 5 413.2 2.4 556.3X
+Partition column - ORC MR 1476 1497 29 10.7 93.9 14.3X
+Both columns - CSV 21051 21724 952 0.7 1338.4 1.0X
+Both columns - Json 12977 13073 137 1.2 825.0 1.6X
+Both columns - Parquet Vectorized: DataPageV1 251 258 6 62.6 16.0 84.2X
+Both columns - Parquet Vectorized: DataPageV2 335 356 14 46.9 21.3 63.1X
+Both columns - Parquet MR: DataPageV1 2966 2969 5 5.3 188.6 7.1X
+Both columns - Parquet MR: DataPageV2 2377 2511 190 6.6 151.1 8.9X
+Both columns - ORC Vectorized 349 357 8 45.1 22.2 60.7X
+Both columns - ORC MR 2418 2458 57 6.5 153.7 8.8X
================================================================================================
String with Nulls Scan
================================================================================================
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (0.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 13555 13582 39 0.8 1292.7 1.0X
-SQL Json 11922 12092 240 0.9 1137.0 1.1X
-SQL Parquet Vectorized: DataPageV1 1695 1696 1 6.2 161.6 8.0X
-SQL Parquet Vectorized: DataPageV2 1890 1891 1 5.5 180.3 7.2X
-SQL Parquet MR: DataPageV1 5013 5091 111 2.1 478.1 2.7X
-SQL Parquet MR: DataPageV2 5450 5465 21 1.9 519.8 2.5X
-ParquetReader Vectorized: DataPageV1 1254 1260 9 8.4 119.6 10.8X
-ParquetReader Vectorized: DataPageV2 1425 1428 5 7.4 135.9 9.5X
-SQL ORC Vectorized 1321 1386 93 7.9 126.0 10.3X
-SQL ORC MR 3816 3861 64 2.7 363.9 3.6X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 11081 11168 122 0.9 1056.8 1.0X
+SQL Json 10522 10543 30 1.0 1003.5 1.1X
+SQL Parquet Vectorized: DataPageV1 1334 1407 103 7.9 127.3 8.3X
+SQL Parquet Vectorized: DataPageV2 1672 1749 109 6.3 159.5 6.6X
+SQL Parquet MR: DataPageV1 4081 4133 73 2.6 389.2 2.7X
+SQL Parquet MR: DataPageV2 3968 3988 29 2.6 378.4 2.8X
+ParquetReader Vectorized: DataPageV1 938 964 23 11.2 89.5 11.8X
+ParquetReader Vectorized: DataPageV2 1057 1065 11 9.9 100.8 10.5X
+SQL ORC Vectorized 1022 1026 6 10.3 97.4 10.8X
+SQL ORC MR 3121 3158 53 3.4 297.6 3.6X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (50.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 9608 9691 117 1.1 916.3 1.0X
-SQL Json 9243 9311 96 1.1 881.5 1.0X
-SQL Parquet Vectorized: DataPageV1 1318 1326 11 8.0 125.7 7.3X
-SQL Parquet Vectorized: DataPageV2 1467 1474 10 7.1 139.9 6.6X
-SQL Parquet MR: DataPageV1 3826 3828 4 2.7 364.9 2.5X
-SQL Parquet MR: DataPageV2 4100 4107 10 2.6 391.0 2.3X
-ParquetReader Vectorized: DataPageV1 1255 1300 64 8.4 119.7 7.7X
-ParquetReader Vectorized: DataPageV2 1482 1509 38 7.1 141.4 6.5X
-SQL ORC Vectorized 1401 1429 39 7.5 133.6 6.9X
-SQL ORC MR 3472 3519 66 3.0 331.1 2.8X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 7851 7863 17 1.3 748.7 1.0X
+SQL Json 7933 7994 86 1.3 756.6 1.0X
+SQL Parquet Vectorized: DataPageV1 965 967 3 10.9 92.0 8.1X
+SQL Parquet Vectorized: DataPageV2 1050 1054 5 10.0 100.1 7.5X
+SQL Parquet MR: DataPageV1 3022 3025 5 3.5 288.2 2.6X
+SQL Parquet MR: DataPageV2 3073 3086 18 3.4 293.1 2.6X
+ParquetReader Vectorized: DataPageV1 876 879 5 12.0 83.6 9.0X
+ParquetReader Vectorized: DataPageV2 953 958 6 11.0 90.8 8.2X
+SQL ORC Vectorized 1151 1154 4 9.1 109.8 6.8X
+SQL ORC MR 2873 2881 12 3.7 274.0 2.7X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (95.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 6357 6388 44 1.6 606.2 1.0X
-SQL Json 5514 5569 78 1.9 525.9 1.2X
-SQL Parquet Vectorized: DataPageV1 356 390 37 29.5 33.9 17.9X
-SQL Parquet Vectorized: DataPageV2 425 446 26 24.7 40.6 14.9X
-SQL Parquet MR: DataPageV1 2377 2388 15 4.4 226.7 2.7X
-SQL Parquet MR: DataPageV2 2170 2239 98 4.8 206.9 2.9X
-ParquetReader Vectorized: DataPageV1 355 362 5 29.5 33.9 17.9X
-ParquetReader Vectorized: DataPageV2 423 440 13 24.8 40.4 15.0X
-SQL ORC Vectorized 448 461 13 23.4 42.7 14.2X
-SQL ORC MR 1924 1932 11 5.4 183.5 3.3X
+SQL CSV 5003 5004 1 2.1 477.1 1.0X
+SQL Json 4778 4825 66 2.2 455.7 1.0X
+SQL Parquet Vectorized: DataPageV1 225 234 6 46.5 21.5 22.2X
+SQL Parquet Vectorized: DataPageV2 249 259 6 42.2 23.7 20.1X
+SQL Parquet MR: DataPageV1 1858 1874 23 5.6 177.2 2.7X
+SQL Parquet MR: DataPageV2 1780 1807 38 5.9 169.8 2.8X
+ParquetReader Vectorized: DataPageV1 227 230 4 46.1 21.7 22.0X
+ParquetReader Vectorized: DataPageV2 253 257 3 41.5 24.1 19.8X
+SQL ORC Vectorized 359 385 20 29.2 34.2 13.9X
+SQL ORC MR 1694 1698 5 6.2 161.6 3.0X
================================================================================================
Single Column Scan From Wide Columns
================================================================================================
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 10 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 2483 2576 131 0.4 2368.1 1.0X
-SQL Json 3806 3821 22 0.3 3629.9 0.7X
-SQL Parquet Vectorized: DataPageV1 50 57 10 21.1 47.5 49.9X
-SQL Parquet Vectorized: DataPageV2 61 69 6 17.2 58.0 40.8X
-SQL Parquet MR: DataPageV1 229 242 14 4.6 218.0 10.9X
-SQL Parquet MR: DataPageV2 217 232 14 4.8 207.1 11.4X
-SQL ORC Vectorized 61 69 8 17.1 58.5 40.5X
-SQL ORC MR 193 208 10 5.4 184.2 12.9X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 2058 2127 97 0.5 1962.8 1.0X
+SQL Json 3291 3316 35 0.3 3138.5 0.6X
+SQL Parquet Vectorized: DataPageV1 43 49 8 24.4 41.0 47.9X
+SQL Parquet Vectorized: DataPageV2 50 56 7 21.1 47.4 41.4X
+SQL Parquet MR: DataPageV1 201 209 8 5.2 191.6 10.2X
+SQL Parquet MR: DataPageV2 187 196 6 5.6 178.2 11.0X
+SQL ORC Vectorized 52 59 8 20.0 50.1 39.2X
+SQL ORC MR 168 177 7 6.3 159.9 12.3X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 50 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 5497 5597 141 0.2 5242.4 1.0X
-SQL Json 15327 15384 81 0.1 14616.5 0.4X
-SQL Parquet Vectorized: DataPageV1 68 75 9 15.4 64.8 80.9X
-SQL Parquet Vectorized: DataPageV2 84 96 11 12.4 80.5 65.2X
-SQL Parquet MR: DataPageV1 259 275 19 4.0 247.2 21.2X
-SQL Parquet MR: DataPageV2 242 254 10 4.3 231.2 22.7X
-SQL ORC Vectorized 82 90 7 12.8 78.3 67.0X
-SQL ORC MR 217 232 12 4.8 206.6 25.4X
-
-OpenJDK 64-Bit Server VM 17.0.2+8-LTS on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 5005 5011 8 0.2 4773.6 1.0X
+SQL Json 13082 13086 7 0.1 12475.8 0.4X
+SQL Parquet Vectorized: DataPageV1 55 62 7 18.9 52.9 90.3X
+SQL Parquet Vectorized: DataPageV2 64 71 9 16.5 60.7 78.7X
+SQL Parquet MR: DataPageV1 218 226 6 4.8 208.1 22.9X
+SQL Parquet MR: DataPageV2 186 204 16 5.6 177.4 26.9X
+SQL ORC Vectorized 67 75 8 15.6 64.2 74.4X
+SQL ORC MR 188 195 7 5.6 179.0 26.7X
+
+OpenJDK 64-Bit Server VM 17.0.4+8-LTS on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 100 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 9101 9254 217 0.1 8679.3 1.0X
-SQL Json 29059 29133 104 0.0 27712.9 0.3X
-SQL Parquet Vectorized: DataPageV1 105 113 7 10.0 99.9 86.9X
-SQL Parquet Vectorized: DataPageV2 117 129 11 9.0 111.5 77.8X
-SQL Parquet MR: DataPageV1 313 323 8 3.4 298.4 29.1X
-SQL Parquet MR: DataPageV2 291 304 13 3.6 277.3 31.3X
-SQL ORC Vectorized 106 119 10 9.9 101.1 85.9X
-SQL ORC MR 269 282 7 3.9 256.3 33.9X
+SQL CSV 8318 8386 97 0.1 7932.5 1.0X
+SQL Json 24615 24712 136 0.0 23475.1 0.3X
+SQL Parquet Vectorized: DataPageV1 85 90 6 12.3 81.3 97.6X
+SQL Parquet Vectorized: DataPageV2 92 97 6 11.5 87.3 90.8X
+SQL Parquet MR: DataPageV1 248 261 8 4.2 236.8 33.5X
+SQL Parquet MR: DataPageV2 233 245 10 4.5 221.8 35.8X
+SQL ORC Vectorized 92 97 6 11.4 87.5 90.6X
+SQL ORC MR 213 226 13 4.9 203.6 39.0X
diff --git a/sql/core/benchmarks/DataSourceReadBenchmark-results.txt b/sql/core/benchmarks/DataSourceReadBenchmark-results.txt
index 2f9ac5d07bd..a4a08bc737a 100644
--- a/sql/core/benchmarks/DataSourceReadBenchmark-results.txt
+++ b/sql/core/benchmarks/DataSourceReadBenchmark-results.txt
@@ -2,430 +2,430 @@
SQL Single Numeric Column Scan
================================================================================================
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 14029 14172 202 1.1 891.9 1.0X
-SQL Json 9040 9105 93 1.7 574.7 1.6X
-SQL Parquet Vectorized: DataPageV1 116 133 15 135.8 7.4 121.1X
-SQL Parquet Vectorized: DataPageV2 101 110 8 155.1 6.4 138.4X
-SQL Parquet MR: DataPageV1 2049 2121 103 7.7 130.3 6.8X
-SQL Parquet MR: DataPageV2 1795 1798 4 8.8 114.1 7.8X
-SQL ORC Vectorized 186 194 7 84.4 11.9 75.2X
-SQL ORC MR 1830 1841 16 8.6 116.3 7.7X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 10369 10492 174 1.5 659.2 1.0X
+SQL Json 7201 7286 120 2.2 457.8 1.4X
+SQL Parquet Vectorized: DataPageV1 125 151 15 126.1 7.9 83.1X
+SQL Parquet Vectorized: DataPageV2 97 110 16 161.9 6.2 106.7X
+SQL Parquet MR: DataPageV1 1800 1815 21 8.7 114.4 5.8X
+SQL Parquet MR: DataPageV2 1597 1602 7 9.8 101.5 6.5X
+SQL ORC Vectorized 176 187 7 89.2 11.2 58.8X
+SQL ORC MR 1591 1594 5 9.9 101.2 6.5X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single BOOLEAN Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 150 154 4 105.0 9.5 1.0X
-ParquetReader Vectorized: DataPageV2 143 145 3 109.6 9.1 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 62 63 3 254.6 3.9 2.4X
-ParquetReader Vectorized -> Row: DataPageV2 55 56 1 288.1 3.5 2.7X
+ParquetReader Vectorized: DataPageV1 128 130 2 122.5 8.2 1.0X
+ParquetReader Vectorized: DataPageV2 120 122 3 130.7 7.6 1.1X
+ParquetReader Vectorized -> Row: DataPageV1 49 50 3 323.8 3.1 2.6X
+ParquetReader Vectorized -> Row: DataPageV2 40 41 1 389.2 2.6 3.2X
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single TINYINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 17060 17175 162 0.9 1084.7 1.0X
-SQL Json 9674 9771 137 1.6 615.0 1.8X
-SQL Parquet Vectorized: DataPageV1 161 168 6 97.7 10.2 106.0X
-SQL Parquet Vectorized: DataPageV2 163 172 10 96.7 10.3 104.9X
-SQL Parquet MR: DataPageV1 2080 2095 22 7.6 132.2 8.2X
-SQL Parquet MR: DataPageV2 1835 1836 1 8.6 116.7 9.3X
-SQL ORC Vectorized 153 156 2 102.6 9.7 111.3X
-SQL ORC MR 1756 1782 37 9.0 111.6 9.7X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 13547 13554 9 1.2 861.3 1.0X
+SQL Json 8705 8714 13 1.8 553.4 1.6X
+SQL Parquet Vectorized: DataPageV1 148 158 10 106.1 9.4 91.4X
+SQL Parquet Vectorized: DataPageV2 147 156 9 106.9 9.4 92.1X
+SQL Parquet MR: DataPageV1 2030 2036 8 7.7 129.1 6.7X
+SQL Parquet MR: DataPageV2 1901 1924 33 8.3 120.9 7.1X
+SQL ORC Vectorized 132 140 4 118.9 8.4 102.4X
+SQL ORC MR 1503 1505 3 10.5 95.5 9.0X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single TINYINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 219 223 4 71.7 13.9 1.0X
-ParquetReader Vectorized: DataPageV2 219 222 3 71.9 13.9 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 223 225 2 70.5 14.2 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 224 226 2 70.3 14.2 1.0X
+ParquetReader Vectorized: DataPageV1 188 190 4 83.9 11.9 1.0X
+ParquetReader Vectorized: DataPageV2 187 190 4 84.1 11.9 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 229 230 1 68.6 14.6 0.8X
+ParquetReader Vectorized -> Row: DataPageV2 229 233 7 68.6 14.6 0.8X
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 17811 18072 369 0.9 1132.4 1.0X
-SQL Json 10172 10183 15 1.5 646.7 1.8X
-SQL Parquet Vectorized: DataPageV1 229 238 10 68.7 14.6 77.8X
-SQL Parquet Vectorized: DataPageV2 250 256 7 63.0 15.9 71.3X
-SQL Parquet MR: DataPageV1 2412 2420 11 6.5 153.4 7.4X
-SQL Parquet MR: DataPageV2 2075 2129 75 7.6 131.9 8.6X
-SQL ORC Vectorized 203 210 9 77.6 12.9 87.9X
-SQL ORC MR 1855 1867 17 8.5 117.9 9.6X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 14262 14305 60 1.1 906.8 1.0X
+SQL Json 9149 9152 4 1.7 581.7 1.6X
+SQL Parquet Vectorized: DataPageV1 214 219 8 73.6 13.6 66.7X
+SQL Parquet Vectorized: DataPageV2 188 199 10 83.5 12.0 75.7X
+SQL Parquet MR: DataPageV1 2255 2258 4 7.0 143.4 6.3X
+SQL Parquet MR: DataPageV2 1932 1945 19 8.1 122.8 7.4X
+SQL ORC Vectorized 183 190 6 85.9 11.6 77.9X
+SQL ORC MR 1641 1645 6 9.6 104.3 8.7X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single SMALLINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 289 294 5 54.4 18.4 1.0X
-ParquetReader Vectorized: DataPageV2 350 354 6 44.9 22.3 0.8X
-ParquetReader Vectorized -> Row: DataPageV1 285 290 7 55.1 18.1 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 347 351 4 45.3 22.1 0.8X
+ParquetReader Vectorized: DataPageV1 250 254 6 62.9 15.9 1.0X
+ParquetReader Vectorized: DataPageV2 269 272 5 58.6 17.1 0.9X
+ParquetReader Vectorized -> Row: DataPageV1 253 255 4 62.3 16.1 1.0X
+ParquetReader Vectorized -> Row: DataPageV2 271 272 1 57.9 17.3 0.9X
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 17952 18152 282 0.9 1141.4 1.0X
-SQL Json 11168 11296 181 1.4 710.0 1.6X
-SQL Parquet Vectorized: DataPageV1 192 203 12 82.0 12.2 93.5X
-SQL Parquet Vectorized: DataPageV2 363 373 10 43.4 23.1 49.5X
-SQL Parquet MR: DataPageV1 2407 2438 44 6.5 153.0 7.5X
-SQL Parquet MR: DataPageV2 2069 2144 106 7.6 131.5 8.7X
-SQL ORC Vectorized 277 284 5 56.7 17.6 64.7X
-SQL ORC MR 2013 2032 28 7.8 128.0 8.9X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 15747 15767 28 1.0 1001.2 1.0X
+SQL Json 9701 9708 10 1.6 616.8 1.6X
+SQL Parquet Vectorized: DataPageV1 173 201 70 91.1 11.0 91.2X
+SQL Parquet Vectorized: DataPageV2 266 272 9 59.2 16.9 59.2X
+SQL Parquet MR: DataPageV1 2252 2278 37 7.0 143.2 7.0X
+SQL Parquet MR: DataPageV2 1988 2033 63 7.9 126.4 7.9X
+SQL ORC Vectorized 254 257 3 62.0 16.1 62.1X
+SQL ORC MR 1679 1687 12 9.4 106.7 9.4X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single INT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 268 273 3 58.7 17.0 1.0X
-ParquetReader Vectorized: DataPageV2 437 446 8 36.0 27.8 0.6X
-ParquetReader Vectorized -> Row: DataPageV1 258 268 8 61.1 16.4 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 429 439 15 36.6 27.3 0.6X
+ParquetReader Vectorized: DataPageV1 235 241 9 67.0 14.9 1.0X
+ParquetReader Vectorized: DataPageV2 345 350 6 45.6 21.9 0.7X
+ParquetReader Vectorized -> Row: DataPageV1 237 239 3 66.5 15.0 1.0X
+ParquetReader Vectorized -> Row: DataPageV2 347 350 5 45.3 22.1 0.7X
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 23444 23655 298 0.7 1490.5 1.0X
-SQL Json 14093 14181 124 1.1 896.0 1.7X
-SQL Parquet Vectorized: DataPageV1 252 262 11 62.3 16.1 92.8X
-SQL Parquet Vectorized: DataPageV2 530 537 10 29.7 33.7 44.2X
-SQL Parquet MR: DataPageV1 2632 2648 23 6.0 167.4 8.9X
-SQL Parquet MR: DataPageV2 2271 2290 26 6.9 144.4 10.3X
-SQL ORC Vectorized 357 362 7 44.0 22.7 65.7X
-SQL ORC MR 2141 2221 113 7.3 136.1 11.0X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 20769 20775 9 0.8 1320.4 1.0X
+SQL Json 12280 12292 16 1.3 780.8 1.7X
+SQL Parquet Vectorized: DataPageV1 235 243 9 66.9 15.0 88.3X
+SQL Parquet Vectorized: DataPageV2 398 404 6 39.5 25.3 52.1X
+SQL Parquet MR: DataPageV1 2354 2361 10 6.7 149.7 8.8X
+SQL Parquet MR: DataPageV2 2085 2113 39 7.5 132.5 10.0X
+SQL ORC Vectorized 323 335 18 48.7 20.5 64.3X
+SQL ORC MR 1805 1806 2 8.7 114.7 11.5X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single BIGINT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 350 396 104 44.9 22.3 1.0X
-ParquetReader Vectorized: DataPageV2 647 736 151 24.3 41.1 0.5X
-ParquetReader Vectorized -> Row: DataPageV1 334 338 7 47.0 21.3 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 624 632 8 25.2 39.7 0.6X
+ParquetReader Vectorized: DataPageV1 306 346 86 51.4 19.5 1.0X
+ParquetReader Vectorized: DataPageV2 483 489 11 32.6 30.7 0.6X
+ParquetReader Vectorized -> Row: DataPageV1 303 308 9 51.9 19.3 1.0X
+ParquetReader Vectorized -> Row: DataPageV2 479 485 11 32.8 30.5 0.6X
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single FLOAT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 19908 20141 329 0.8 1265.7 1.0X
-SQL Json 13175 13196 30 1.2 837.6 1.5X
-SQL Parquet Vectorized: DataPageV1 167 174 7 94.0 10.6 118.9X
-SQL Parquet Vectorized: DataPageV2 167 171 6 94.2 10.6 119.2X
-SQL Parquet MR: DataPageV1 2401 2448 66 6.6 152.6 8.3X
-SQL Parquet MR: DataPageV2 2211 2238 38 7.1 140.6 9.0X
-SQL ORC Vectorized 401 404 3 39.2 25.5 49.7X
-SQL ORC MR 2046 2047 2 7.7 130.1 9.7X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 15664 15731 95 1.0 995.9 1.0X
+SQL Json 11728 11733 6 1.3 745.7 1.3X
+SQL Parquet Vectorized: DataPageV1 155 162 8 101.2 9.9 100.8X
+SQL Parquet Vectorized: DataPageV2 154 158 5 102.1 9.8 101.6X
+SQL Parquet MR: DataPageV1 2221 2249 40 7.1 141.2 7.1X
+SQL Parquet MR: DataPageV2 2095 2101 8 7.5 133.2 7.5X
+SQL ORC Vectorized 362 366 3 43.4 23.0 43.2X
+SQL ORC MR 1787 1788 1 8.8 113.6 8.8X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single FLOAT Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 280 287 6 56.2 17.8 1.0X
-ParquetReader Vectorized: DataPageV2 286 310 54 55.0 18.2 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 277 281 4 56.9 17.6 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 274 278 3 57.4 17.4 1.0X
+ParquetReader Vectorized: DataPageV1 254 260 7 61.9 16.2 1.0X
+ParquetReader Vectorized: DataPageV2 254 259 9 61.9 16.2 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 247 251 7 63.8 15.7 1.0X
+ParquetReader Vectorized -> Row: DataPageV2 246 249 5 63.9 15.7 1.0X
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single DOUBLE Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 26970 27069 139 0.6 1714.7 1.0X
-SQL Json 18898 18975 109 0.8 1201.5 1.4X
-SQL Parquet Vectorized: DataPageV1 239 245 7 65.8 15.2 112.9X
-SQL Parquet Vectorized: DataPageV2 245 250 7 64.3 15.6 110.2X
-SQL Parquet MR: DataPageV1 2726 2740 19 5.8 173.3 9.9X
-SQL Parquet MR: DataPageV2 2409 2410 2 6.5 153.1 11.2X
-SQL ORC Vectorized 469 471 1 33.5 29.8 57.5X
-SQL ORC MR 2143 2143 0 7.3 136.2 12.6X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 20816 20835 27 0.8 1323.4 1.0X
+SQL Json 15792 15801 14 1.0 1004.0 1.3X
+SQL Parquet Vectorized: DataPageV1 233 242 12 67.5 14.8 89.3X
+SQL Parquet Vectorized: DataPageV2 235 248 23 67.0 14.9 88.7X
+SQL Parquet MR: DataPageV1 2360 2365 7 6.7 150.0 8.8X
+SQL Parquet MR: DataPageV2 2247 2251 6 7.0 142.9 9.3X
+SQL ORC Vectorized 440 443 2 35.8 28.0 47.3X
+SQL ORC MR 1884 1887 4 8.3 119.8 11.0X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Parquet Reader Single DOUBLE Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------
-ParquetReader Vectorized: DataPageV1 345 358 13 45.5 22.0 1.0X
-ParquetReader Vectorized: DataPageV2 352 362 14 44.7 22.4 1.0X
-ParquetReader Vectorized -> Row: DataPageV1 342 351 10 46.0 21.7 1.0X
-ParquetReader Vectorized -> Row: DataPageV2 352 359 7 44.7 22.4 1.0X
+ParquetReader Vectorized: DataPageV1 320 331 16 49.2 20.3 1.0X
+ParquetReader Vectorized: DataPageV2 323 332 14 48.7 20.5 1.0X
+ParquetReader Vectorized -> Row: DataPageV1 309 314 7 50.9 19.6 1.0X
+ParquetReader Vectorized -> Row: DataPageV2 311 315 7 50.5 19.8 1.0X
================================================================================================
SQL Single Numeric Column Scan in Struct
================================================================================================
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single TINYINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 2593 2598 6 6.1 164.9 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 2590 2595 8 6.1 164.7 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 360 370 14 43.7 22.9 7.2X
-SQL Parquet MR: DataPageV1 2701 2708 10 5.8 171.7 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3279 3281 2 4.8 208.5 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 355 360 7 44.3 22.6 7.3X
-SQL Parquet MR: DataPageV2 2613 2659 65 6.0 166.1 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3099 3157 83 5.1 197.0 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 357 362 4 44.1 22.7 7.3X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2239 2245 9 7.0 142.3 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2237 2241 6 7.0 142.2 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 320 330 9 49.1 20.4 7.0X
+SQL Parquet MR: DataPageV1 2351 2355 5 6.7 149.5 1.0X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 2780 2780 1 5.7 176.7 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 308 313 3 51.1 19.6 7.3X
+SQL Parquet MR: DataPageV2 2234 2238 6 7.0 142.0 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2625 2626 1 6.0 166.9 0.9X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 306 312 6 51.5 19.4 7.3X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single SMALLINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 2717 2736 26 5.8 172.8 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 2712 2724 17 5.8 172.4 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 371 376 4 42.4 23.6 7.3X
-SQL Parquet MR: DataPageV1 2947 2994 67 5.3 187.3 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3518 3557 55 4.5 223.7 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 340 358 17 46.2 21.6 8.0X
-SQL Parquet MR: DataPageV2 2693 2787 132 5.8 171.2 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3249 3274 36 4.8 206.5 0.8X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 523 528 7 30.1 33.3 5.2X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2267 2276 13 6.9 144.1 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2290 2300 13 6.9 145.6 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 356 383 21 44.2 22.6 6.4X
+SQL Parquet MR: DataPageV1 2567 2590 32 6.1 163.2 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3067 3071 5 5.1 195.0 0.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 311 319 9 50.5 19.8 7.3X
+SQL Parquet MR: DataPageV2 2346 2355 13 6.7 149.1 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2798 2801 4 5.6 177.9 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 428 429 2 36.8 27.2 5.3X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single INT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 2720 2748 39 5.8 173.0 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 2772 2792 28 5.7 176.2 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 467 472 5 33.7 29.7 5.8X
-SQL Parquet MR: DataPageV1 2879 2901 30 5.5 183.1 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3536 3617 113 4.4 224.8 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 310 328 19 50.7 19.7 8.8X
-SQL Parquet MR: DataPageV2 2602 2637 50 6.0 165.4 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3065 3119 78 5.1 194.8 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 569 584 19 27.6 36.2 4.8X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2327 2336 13 6.8 148.0 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2309 2324 21 6.8 146.8 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 415 453 24 37.9 26.4 5.6X
+SQL Parquet MR: DataPageV1 2617 2625 11 6.0 166.4 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3115 3119 6 5.0 198.0 0.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 298 315 15 52.8 18.9 7.8X
+SQL Parquet MR: DataPageV2 2403 2421 26 6.5 152.8 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2882 2885 5 5.5 183.2 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 471 479 6 33.4 30.0 4.9X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single BIGINT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 2854 2871 24 5.5 181.5 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 2714 2737 32 5.8 172.6 1.1X
-SQL ORC Vectorized (Nested Column Enabled) 502 508 5 31.3 31.9 5.7X
-SQL Parquet MR: DataPageV1 2929 2992 90 5.4 186.2 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3563 3572 14 4.4 226.5 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 400 431 22 39.3 25.5 7.1X
-SQL Parquet MR: DataPageV2 2827 2866 55 5.6 179.7 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3314 3330 22 4.7 210.7 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 739 757 16 21.3 47.0 3.9X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2394 2431 53 6.6 152.2 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2405 2405 0 6.5 152.9 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 507 515 6 31.0 32.2 4.7X
+SQL Parquet MR: DataPageV1 2748 2774 37 5.7 174.7 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3233 3234 1 4.9 205.6 0.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 378 385 12 41.6 24.0 6.3X
+SQL Parquet MR: DataPageV2 2552 2553 1 6.2 162.3 0.9X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2940 2941 0 5.3 186.9 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 588 598 13 26.8 37.4 4.1X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single FLOAT Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 2919 2926 9 5.4 185.6 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 2855 2875 28 5.5 181.5 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 613 616 4 25.7 39.0 4.8X
-SQL Parquet MR: DataPageV1 2859 2869 13 5.5 181.8 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3419 3422 4 4.6 217.4 0.9X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 322 326 8 48.9 20.4 9.1X
-SQL Parquet MR: DataPageV2 2738 2748 14 5.7 174.1 1.1X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3266 3276 14 4.8 207.7 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 314 329 15 50.0 20.0 9.3X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL ORC MR 2408 2431 33 6.5 153.1 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2430 2452 31 6.5 154.5 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 539 545 8 29.2 34.3 4.5X
+SQL Parquet MR: DataPageV1 2566 2575 13 6.1 163.1 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3053 3069 22 5.2 194.1 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 294 312 8 53.5 18.7 8.2X
+SQL Parquet MR: DataPageV2 2457 2460 4 6.4 156.2 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 2894 2901 11 5.4 184.0 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 293 299 9 53.6 18.6 8.2X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Single DOUBLE Column Scan in Struct: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 2920 2927 10 5.4 185.6 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 2930 2948 25 5.4 186.3 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 643 651 7 24.4 40.9 4.5X
-SQL Parquet MR: DataPageV1 3053 3067 20 5.2 194.1 1.0X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3760 3774 20 4.2 239.0 0.8X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 405 427 16 38.9 25.7 7.2X
-SQL Parquet MR: DataPageV2 2986 2995 12 5.3 189.9 1.0X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3422 3486 91 4.6 217.5 0.9X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 392 402 10 40.1 24.9 7.4X
+SQL ORC MR 2484 2625 198 6.3 158.0 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 2485 2487 3 6.3 158.0 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 601 604 4 26.2 38.2 4.1X
+SQL Parquet MR: DataPageV1 2760 2762 2 5.7 175.5 0.9X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 3227 3232 7 4.9 205.2 0.8X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 388 406 21 40.5 24.7 6.4X
+SQL Parquet MR: DataPageV2 2589 2616 38 6.1 164.6 1.0X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 3003 3030 37 5.2 190.9 0.8X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 369 382 16 42.6 23.5 6.7X
================================================================================================
SQL Nested Column Scan
================================================================================================
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
SQL Nested Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
-------------------------------------------------------------------------------------------------------------------------------------------
-SQL ORC MR 25774 26314 362 0.0 24580.2 1.0X
-SQL ORC Vectorized (Nested Column Disabled) 25353 25788 244 0.0 24178.7 1.0X
-SQL ORC Vectorized (Nested Column Enabled) 9166 9243 42 0.1 8741.0 2.8X
-SQL Parquet MR: DataPageV1 15232 15456 196 0.1 14526.0 1.7X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 16231 16411 144 0.1 15479.3 1.6X
-SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 10501 10976 612 0.1 10014.6 2.5X
-SQL Parquet MR: DataPageV2 18227 18605 272 0.1 17382.8 1.4X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 19216 19499 224 0.1 18325.7 1.3X
-SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 8328 8714 269 0.1 7942.4 3.1X
+SQL ORC MR 24717 25116 254 0.0 23572.3 1.0X
+SQL ORC Vectorized (Nested Column Disabled) 24773 25074 230 0.0 23625.5 1.0X
+SQL ORC Vectorized (Nested Column Enabled) 8992 9290 230 0.1 8575.9 2.7X
+SQL Parquet MR: DataPageV1 14232 15261 863 0.1 13572.8 1.7X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Disabled) 16288 18150 NaN 0.1 15533.3 1.5X
+SQL Parquet Vectorized: DataPageV1 (Nested Column Enabled) 12868 17513 NaN 0.1 12272.3 1.9X
+SQL Parquet MR: DataPageV2 17112 17361 344 0.1 16319.1 1.4X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Disabled) 17886 17974 126 0.1 17057.0 1.4X
+SQL Parquet Vectorized: DataPageV2 (Nested Column Enabled) 7769 7843 70 0.1 7409.1 3.2X
================================================================================================
Int and String Scan
================================================================================================
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Int and String Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 17749 18067 450 0.6 1692.7 1.0X
-SQL Json 14375 14442 95 0.7 1370.9 1.2X
-SQL Parquet Vectorized: DataPageV1 2413 2424 15 4.3 230.1 7.4X
-SQL Parquet Vectorized: DataPageV2 2851 2890 56 3.7 271.9 6.2X
-SQL Parquet MR: DataPageV1 4649 4664 20 2.3 443.4 3.8X
-SQL Parquet MR: DataPageV2 4450 4462 17 2.4 424.3 4.0X
-SQL ORC Vectorized 2255 2264 13 4.6 215.1 7.9X
-SQL ORC MR 4168 4174 9 2.5 397.5 4.3X
+SQL CSV 14413 14445 45 0.7 1374.5 1.0X
+SQL Json 11852 11887 50 0.9 1130.3 1.2X
+SQL Parquet Vectorized: DataPageV1 2087 2091 6 5.0 199.0 6.9X
+SQL Parquet Vectorized: DataPageV2 2361 2379 24 4.4 225.2 6.1X
+SQL Parquet MR: DataPageV1 4189 4209 28 2.5 399.5 3.4X
+SQL Parquet MR: DataPageV2 4302 4310 12 2.4 410.2 3.4X
+SQL ORC Vectorized 2089 2097 10 5.0 199.3 6.9X
+SQL ORC MR 3756 3763 10 2.8 358.2 3.8X
================================================================================================
Repeated String Scan
================================================================================================
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Repeated String: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 9351 9365 19 1.1 891.8 1.0X
-SQL Json 8466 8492 37 1.2 807.4 1.1X
-SQL Parquet Vectorized: DataPageV1 808 819 10 13.0 77.1 11.6X
-SQL Parquet Vectorized: DataPageV2 779 790 16 13.5 74.3 12.0X
-SQL Parquet MR: DataPageV1 1854 1856 3 5.7 176.8 5.0X
-SQL Parquet MR: DataPageV2 1773 1777 5 5.9 169.1 5.3X
-SQL ORC Vectorized 497 511 11 21.1 47.4 18.8X
-SQL ORC MR 2078 2100 31 5.0 198.2 4.5X
+SQL CSV 7583 7633 71 1.4 723.1 1.0X
+SQL Json 7036 7060 33 1.5 671.0 1.1X
+SQL Parquet Vectorized: DataPageV1 706 713 8 14.9 67.3 10.7X
+SQL Parquet Vectorized: DataPageV2 700 704 7 15.0 66.8 10.8X
+SQL Parquet MR: DataPageV1 1736 1741 7 6.0 165.5 4.4X
+SQL Parquet MR: DataPageV2 1663 1669 9 6.3 158.6 4.6X
+SQL ORC Vectorized 448 457 8 23.4 42.7 16.9X
+SQL ORC MR 1861 1868 11 5.6 177.4 4.1X
================================================================================================
Partitioned Table Scan
================================================================================================
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Partitioned Table: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
---------------------------------------------------------------------------------------------------------------------------------
-Data column - CSV 26283 26998 1010 0.6 1671.0 1.0X
-Data column - Json 15661 15688 38 1.0 995.7 1.7X
-Data column - Parquet Vectorized: DataPageV1 246 257 6 64.0 15.6 107.0X
-Data column - Parquet Vectorized: DataPageV2 592 611 17 26.6 37.6 44.4X
-Data column - Parquet MR: DataPageV1 3018 3033 21 5.2 191.9 8.7X
-Data column - Parquet MR: DataPageV2 2762 2765 4 5.7 175.6 9.5X
-Data column - ORC Vectorized 349 353 4 45.1 22.2 75.4X
-Data column - ORC MR 2544 2545 1 6.2 161.8 10.3X
-Partition column - CSV 7570 7620 70 2.1 481.3 3.5X
-Partition column - Json 11928 12048 169 1.3 758.4 2.2X
-Partition column - Parquet Vectorized: DataPageV1 61 66 5 259.3 3.9 433.3X
-Partition column - Parquet Vectorized: DataPageV2 60 65 5 260.7 3.8 435.7X
-Partition column - Parquet MR: DataPageV1 1385 1396 16 11.4 88.0 19.0X
-Partition column - Parquet MR: DataPageV2 1429 1437 11 11.0 90.8 18.4X
-Partition column - ORC Vectorized 63 68 4 250.2 4.0 418.1X
-Partition column - ORC MR 1595 1605 13 9.9 101.4 16.5X
-Both columns - CSV 25899 26018 168 0.6 1646.6 1.0X
-Both columns - Json 14667 14685 26 1.1 932.5 1.8X
-Both columns - Parquet Vectorized: DataPageV1 288 295 10 54.7 18.3 91.4X
-Both columns - Parquet Vectorized: DataPageV2 621 624 3 25.3 39.5 42.3X
-Both columns - Parquet MR: DataPageV1 2905 2945 57 5.4 184.7 9.0X
-Both columns - Parquet MR: DataPageV2 2772 2790 25 5.7 176.2 9.5X
-Both columns - ORC Vectorized 401 405 5 39.3 25.5 65.6X
-Both columns - ORC MR 2601 2610 12 6.0 165.4 10.1X
+Data column - CSV 19162 19201 55 0.8 1218.3 1.0X
+Data column - Json 12407 12428 30 1.3 788.8 1.5X
+Data column - Parquet Vectorized: DataPageV1 233 242 8 67.6 14.8 82.4X
+Data column - Parquet Vectorized: DataPageV2 449 460 10 35.0 28.6 42.7X
+Data column - Parquet MR: DataPageV1 2754 2790 51 5.7 175.1 7.0X
+Data column - Parquet MR: DataPageV2 2540 2541 2 6.2 161.5 7.5X
+Data column - ORC Vectorized 321 330 5 49.0 20.4 59.7X
+Data column - ORC MR 2146 2150 6 7.3 136.4 8.9X
+Partition column - CSV 6364 6365 1 2.5 404.6 3.0X
+Partition column - Json 9733 9736 4 1.6 618.8 2.0X
+Partition column - Parquet Vectorized: DataPageV1 49 55 14 318.5 3.1 388.0X
+Partition column - Parquet Vectorized: DataPageV2 50 56 12 313.4 3.2 381.8X
+Partition column - Parquet MR: DataPageV1 1353 1355 3 11.6 86.0 14.2X
+Partition column - Parquet MR: DataPageV2 1351 1370 27 11.6 85.9 14.2X
+Partition column - ORC Vectorized 50 58 14 313.6 3.2 382.1X
+Partition column - ORC MR 1349 1351 3 11.7 85.7 14.2X
+Both columns - CSV 19854 19860 8 0.8 1262.3 1.0X
+Both columns - Json 13054 13070 23 1.2 830.0 1.5X
+Both columns - Parquet Vectorized: DataPageV1 276 283 9 57.0 17.6 69.4X
+Both columns - Parquet Vectorized: DataPageV2 490 498 8 32.1 31.2 39.1X
+Both columns - Parquet MR: DataPageV1 2822 2832 15 5.6 179.4 6.8X
+Both columns - Parquet MR: DataPageV2 2590 2595 6 6.1 164.7 7.4X
+Both columns - ORC Vectorized 361 364 3 43.6 22.9 53.1X
+Both columns - ORC MR 2295 2295 1 6.9 145.9 8.4X
================================================================================================
String with Nulls Scan
================================================================================================
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (0.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 11883 11921 54 0.9 1133.2 1.0X
-SQL Json 12458 12492 48 0.8 1188.1 1.0X
-SQL Parquet Vectorized: DataPageV1 1483 1486 4 7.1 141.4 8.0X
-SQL Parquet Vectorized: DataPageV2 1958 1987 41 5.4 186.7 6.1X
-SQL Parquet MR: DataPageV1 3858 3865 11 2.7 367.9 3.1X
-SQL Parquet MR: DataPageV2 4526 4543 24 2.3 431.6 2.6X
-ParquetReader Vectorized: DataPageV1 1085 1094 13 9.7 103.4 11.0X
-ParquetReader Vectorized: DataPageV2 1620 1623 4 6.5 154.5 7.3X
-SQL ORC Vectorized 974 1007 46 10.8 92.9 12.2X
-SQL ORC MR 3499 3501 2 3.0 333.7 3.4X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 9386 9424 54 1.1 895.1 1.0X
+SQL Json 10896 10918 31 1.0 1039.1 0.9X
+SQL Parquet Vectorized: DataPageV1 1296 1311 20 8.1 123.6 7.2X
+SQL Parquet Vectorized: DataPageV2 1735 1736 0 6.0 165.5 5.4X
+SQL Parquet MR: DataPageV1 3508 3510 2 3.0 334.5 2.7X
+SQL Parquet MR: DataPageV2 4558 4561 4 2.3 434.7 2.1X
+ParquetReader Vectorized: DataPageV1 962 971 15 10.9 91.7 9.8X
+ParquetReader Vectorized: DataPageV2 1363 1365 3 7.7 130.0 6.9X
+SQL ORC Vectorized 1031 1044 19 10.2 98.3 9.1X
+SQL ORC MR 3233 3239 8 3.2 308.3 2.9X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (50.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 8465 8604 197 1.2 807.3 1.0X
-SQL Json 9376 9382 8 1.1 894.2 0.9X
-SQL Parquet Vectorized: DataPageV1 1226 1234 11 8.6 116.9 6.9X
-SQL Parquet Vectorized: DataPageV2 1466 1474 11 7.2 139.8 5.8X
-SQL Parquet MR: DataPageV1 2896 2934 54 3.6 276.2 2.9X
-SQL Parquet MR: DataPageV2 3407 3432 35 3.1 324.9 2.5X
-ParquetReader Vectorized: DataPageV1 1132 1137 8 9.3 107.9 7.5X
-ParquetReader Vectorized: DataPageV2 1471 1489 26 7.1 140.3 5.8X
-SQL ORC Vectorized 1344 1351 10 7.8 128.1 6.3X
-SQL ORC MR 3533 3540 9 3.0 336.9 2.4X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 6765 6772 11 1.6 645.1 1.0X
+SQL Json 8040 8048 11 1.3 766.8 0.8X
+SQL Parquet Vectorized: DataPageV1 1046 1063 24 10.0 99.7 6.5X
+SQL Parquet Vectorized: DataPageV2 1272 1280 11 8.2 121.3 5.3X
+SQL Parquet MR: DataPageV1 2740 2747 10 3.8 261.3 2.5X
+SQL Parquet MR: DataPageV2 3210 3212 3 3.3 306.2 2.1X
+ParquetReader Vectorized: DataPageV1 939 946 9 11.2 89.6 7.2X
+ParquetReader Vectorized: DataPageV2 1190 1197 9 8.8 113.5 5.7X
+SQL ORC Vectorized 1138 1142 5 9.2 108.6 5.9X
+SQL ORC MR 2934 2937 4 3.6 279.8 2.3X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
String with Nulls Scan (95.0%): Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 6105 6192 123 1.7 582.2 1.0X
-SQL Json 5335 5365 42 2.0 508.8 1.1X
-SQL Parquet Vectorized: DataPageV1 285 290 5 36.7 27.2 21.4X
-SQL Parquet Vectorized: DataPageV2 342 347 5 30.7 32.6 17.9X
-SQL Parquet MR: DataPageV1 2008 2015 9 5.2 191.5 3.0X
-SQL Parquet MR: DataPageV2 1904 1924 29 5.5 181.6 3.2X
-ParquetReader Vectorized: DataPageV1 314 319 4 33.4 30.0 19.4X
-ParquetReader Vectorized: DataPageV2 370 374 4 28.3 35.3 16.5X
-SQL ORC Vectorized 460 468 7 22.8 43.9 13.3X
-SQL ORC MR 2064 2066 2 5.1 196.9 3.0X
+SQL CSV 4467 4472 7 2.3 426.0 1.0X
+SQL Json 4640 4640 1 2.3 442.5 1.0X
+SQL Parquet Vectorized: DataPageV1 230 237 12 45.6 21.9 19.4X
+SQL Parquet Vectorized: DataPageV2 279 287 6 37.6 26.6 16.0X
+SQL Parquet MR: DataPageV1 1822 1829 9 5.8 173.8 2.5X
+SQL Parquet MR: DataPageV2 1774 1779 7 5.9 169.2 2.5X
+ParquetReader Vectorized: DataPageV1 240 244 5 43.7 22.9 18.6X
+ParquetReader Vectorized: DataPageV2 290 293 3 36.2 27.6 15.4X
+SQL ORC Vectorized 363 366 2 28.9 34.6 12.3X
+SQL ORC MR 1544 1547 3 6.8 147.3 2.9X
================================================================================================
Single Column Scan From Wide Columns
================================================================================================
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 10 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 2674 2678 7 0.4 2549.8 1.0X
-SQL Json 3802 3818 23 0.3 3626.0 0.7X
-SQL Parquet Vectorized: DataPageV1 49 54 7 21.2 47.1 54.1X
-SQL Parquet Vectorized: DataPageV2 72 78 8 14.5 69.0 37.0X
-SQL Parquet MR: DataPageV1 221 233 8 4.7 211.1 12.1X
-SQL Parquet MR: DataPageV2 198 206 4 5.3 188.5 13.5X
-SQL ORC Vectorized 59 66 11 17.6 56.7 45.0X
-SQL ORC MR 174 181 5 6.0 165.8 15.4X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 2206 2209 4 0.5 2103.6 1.0X
+SQL Json 2873 2883 15 0.4 2739.5 0.8X
+SQL Parquet Vectorized: DataPageV1 48 52 7 21.7 46.1 45.6X
+SQL Parquet Vectorized: DataPageV2 64 68 9 16.4 60.9 34.5X
+SQL Parquet MR: DataPageV1 204 209 4 5.1 194.4 10.8X
+SQL Parquet MR: DataPageV2 188 194 7 5.6 179.6 11.7X
+SQL ORC Vectorized 55 59 8 19.1 52.4 40.1X
+SQL ORC MR 161 164 2 6.5 153.3 13.7X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 50 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 6572 6614 59 0.2 6267.8 1.0X
-SQL Json 15786 15817 43 0.1 15055.2 0.4X
-SQL Parquet Vectorized: DataPageV1 73 80 11 14.4 69.6 90.1X
-SQL Parquet Vectorized: DataPageV2 94 102 10 11.2 89.3 70.2X
-SQL Parquet MR: DataPageV1 251 258 6 4.2 238.9 26.2X
-SQL Parquet MR: DataPageV2 225 230 4 4.7 214.5 29.2X
-SQL ORC Vectorized 76 82 7 13.8 72.5 86.4X
-SQL ORC MR 196 200 3 5.4 186.6 33.6X
-
-OpenJDK 64-Bit Server VM 1.8.0_322-b06 on Linux 5.13.0-1021-azure
-Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
+SQL CSV 5378 5391 18 0.2 5128.7 1.0X
+SQL Json 11196 11225 42 0.1 10676.9 0.5X
+SQL Parquet Vectorized: DataPageV1 64 72 12 16.5 60.6 84.6X
+SQL Parquet Vectorized: DataPageV2 79 85 8 13.3 75.0 68.3X
+SQL Parquet MR: DataPageV1 223 230 10 4.7 212.6 24.1X
+SQL Parquet MR: DataPageV2 205 214 11 5.1 195.7 26.2X
+SQL ORC Vectorized 68 76 12 15.4 65.1 78.7X
+SQL ORC MR 178 185 6 5.9 169.4 30.3X
+
+OpenJDK 64-Bit Server VM 1.8.0_342-b07 on Linux 5.15.0-1014-azure
+Intel(R) Xeon(R) Platinum 8272CL CPU @ 2.60GHz
Single Column Scan from 100 columns: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
-SQL CSV 11953 11983 42 0.1 11399.4 1.0X
-SQL Json 30922 31434 725 0.0 29489.2 0.4X
-SQL Parquet Vectorized: DataPageV1 107 112 7 9.8 102.5 111.2X
-SQL Parquet Vectorized: DataPageV2 129 135 4 8.1 123.4 92.4X
-SQL Parquet MR: DataPageV1 288 294 6 3.6 275.0 41.5X
-SQL Parquet MR: DataPageV2 267 276 10 3.9 254.2 44.8X
-SQL ORC Vectorized 107 110 7 9.8 102.1 111.7X
-SQL ORC MR 228 238 10 4.6 217.4 52.4X
+SQL CSV 9249 9256 10 0.1 8820.1 1.0X
+SQL Json 21816 21866 70 0.0 20805.6 0.4X
+SQL Parquet Vectorized: DataPageV1 92 100 8 11.3 88.1 100.1X
+SQL Parquet Vectorized: DataPageV2 112 119 7 9.3 107.3 82.2X
+SQL Parquet MR: DataPageV1 258 262 6 4.1 245.8 35.9X
+SQL Parquet MR: DataPageV2 246 251 7 4.3 234.2 37.7X
+SQL ORC Vectorized 92 99 10 11.3 88.2 100.1X
+SQL ORC MR 198 208 12 5.3 188.7 46.7X
diff --git a/sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/VectorizedDeltaBinaryPackedReader.java b/sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/VectorizedDeltaBinaryPackedReader.java
index 9c6596aa1b5..d6d7d8ea2b0 100644
--- a/sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/VectorizedDeltaBinaryPackedReader.java
+++ b/sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/VectorizedDeltaBinaryPackedReader.java
@@ -300,7 +300,8 @@ public class VectorizedDeltaBinaryPackedReader extends VectorizedReaderBase {
bitWidths[currentMiniBlock]);
for (int j = 0; j < miniBlockSizeInValues; j += 8) {
ByteBuffer buffer = in.slice(packer.getBitWidth());
- packer.unpack8Values(buffer, buffer.position(), unpackedValuesBuffer, j);
+ packer.unpack8Values(buffer.array(),
+ buffer.arrayOffset() + buffer.position(), unpackedValuesBuffer, j);
}
remainingInMiniBlock = miniBlockSizeInValues;
currentMiniBlock++;
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@spark.apache.org
For additional commands, e-mail: commits-help@spark.apache.org