ODIn/spark-instrumented-optimizer
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity
spark-instrumented-optimizer/sql/core/benchmarks/DateTimeRebaseBenchmark-jdk11-results.txt
HyukjinKwon ebf01ec3c1 [SPARK-34950][TESTS] Update benchmark results to the ones created by GitHub Actions machines 
### What changes were proposed in this pull request?

https://github.com/apache/spark/pull/32015 added a way to run benchmarks much more easily in the same GitHub Actions build. This PR updates the benchmark results by using the way.

**NOTE** that looks like GitHub Actions use four types of CPU given my observations:

- Intel(R) Xeon(R) Platinum 8171M CPU  2.60GHz
- Intel(R) Xeon(R) CPU E5-2673 v4  2.30GHz
- Intel(R) Xeon(R) CPU E5-2673 v3  2.40GHz
- Intel(R) Xeon(R) Platinum 8272CL CPU  2.60GHz

Given my quick research, seems like they perform roughly similarly:

![Screen Shot 2021-04-03 at 9 31 23 PM](https://user-images.githubusercontent.com/6477701/113478478-f4b57b80-94c3-11eb-9047-f81ca8c59672.png)

I couldn't find enough information about Intel(R) Xeon(R) Platinum 8272CL CPU  2.60GHz but the performance seems roughly similar given the numbers.

So shouldn't be a big deal especially given that this way is much easier, encourages contributors to run more and guarantee the same number of cores and same memory with the same softwares.

### Why are the changes needed?

To have a base line of the benchmarks accordingly.

### Does this PR introduce _any_ user-facing change?

No, dev-only.

### How was this patch tested?

It was generated from:

- [Run benchmarks: * (JDK 11)](https://github.com/HyukjinKwon/spark/actions/runs/713575465)
- [Run benchmarks: * (JDK 8)](https://github.com/HyukjinKwon/spark/actions/runs/713154337)

Closes #32044 from HyukjinKwon/SPARK-34950.

Authored-by: HyukjinKwon <gurwls223@apache.org>
Signed-off-by: Max Gekk <max.gekk@gmail.com>
2021-04-03 23:02:56 +03:00

155 lines
14 KiB
Plaintext
Raw Blame History

================================================================================================
Rebasing dates/timestamps in Parquet datasource
================================================================================================
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Save DATE to parquet: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1582, noop 24957 24957 0 4.0 249.6 1.0X
before 1582, noop 13309 13309 0 7.5 133.1 1.9X
after 1582, rebase EXCEPTION 38895 38895 0 2.6 388.9 0.6X
after 1582, rebase LEGACY 38477 38477 0 2.6 384.8 0.6X
after 1582, rebase CORRECTED 39442 39442 0 2.5 394.4 0.6X
before 1582, rebase LEGACY 31573 31573 0 3.2 315.7 0.8X
before 1582, rebase CORRECTED 28340 28340 0 3.5 283.4 0.9X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Load DATE from parquet: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1582, vec off, rebase EXCEPTION 18452 18602 146 5.4 184.5 1.0X
after 1582, vec off, rebase LEGACY 18570 18609 36 5.4 185.7 1.0X
after 1582, vec off, rebase CORRECTED 18432 18646 191 5.4 184.3 1.0X
after 1582, vec on, rebase EXCEPTION 4473 4581 120 22.4 44.7 4.1X
after 1582, vec on, rebase LEGACY 4449 4548 87 22.5 44.5 4.1X
after 1582, vec on, rebase CORRECTED 4409 4435 32 22.7 44.1 4.2X
before 1582, vec off, rebase LEGACY 19835 19969 116 5.0 198.4 0.9X
before 1582, vec off, rebase CORRECTED 19035 19054 29 5.3 190.4 1.0X
before 1582, vec on, rebase LEGACY 5843 5929 91 17.1 58.4 3.2X
before 1582, vec on, rebase CORRECTED 4274 4347 74 23.4 42.7 4.3X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Save TIMESTAMP_INT96 to parquet: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1900, noop 3409 3409 0 29.3 34.1 1.0X
before 1900, noop 3664 3664 0 27.3 36.6 0.9X
after 1900, rebase EXCEPTION 33008 33008 0 3.0 330.1 0.1X
after 1900, rebase LEGACY 32753 32753 0 3.1 327.5 0.1X
after 1900, rebase CORRECTED 34472 34472 0 2.9 344.7 0.1X
before 1900, rebase LEGACY 38110 38110 0 2.6 381.1 0.1X
before 1900, rebase CORRECTED 35285 35285 0 2.8 352.9 0.1X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Load TIMESTAMP_INT96 from parquet: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1900, vec off, rebase EXCEPTION 27084 27321 216 3.7 270.8 1.0X
after 1900, vec off, rebase LEGACY 26556 27280 630 3.8 265.6 1.0X
after 1900, vec off, rebase CORRECTED 26710 27071 319 3.7 267.1 1.0X
after 1900, vec on, rebase EXCEPTION 8552 8693 168 11.7 85.5 3.2X
after 1900, vec on, rebase LEGACY 8853 8866 12 11.3 88.5 3.1X
after 1900, vec on, rebase CORRECTED 8161 8327 221 12.3 81.6 3.3X
before 1900, vec off, rebase LEGACY 30156 30951 702 3.3 301.6 0.9X
before 1900, vec off, rebase CORRECTED 27258 27418 139 3.7 272.6 1.0X
before 1900, vec on, rebase LEGACY 12023 12086 89 8.3 120.2 2.3X
before 1900, vec on, rebase CORRECTED 8546 8554 6 11.7 85.5 3.2X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Save TIMESTAMP_MICROS to parquet: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1900, noop 3765 3765 0 26.6 37.7 1.0X
before 1900, noop 3563 3563 0 28.1 35.6 1.1X
after 1900, rebase EXCEPTION 21884 21884 0 4.6 218.8 0.2X
after 1900, rebase LEGACY 23126 23126 0 4.3 231.3 0.2X
after 1900, rebase CORRECTED 24592 24592 0 4.1 245.9 0.2X
before 1900, rebase LEGACY 27854 27854 0 3.6 278.5 0.1X
before 1900, rebase CORRECTED 24708 24708 0 4.0 247.1 0.2X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Load TIMESTAMP_MICROS from parquet: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1900, vec off, rebase EXCEPTION 21394 21679 253 4.7 213.9 1.0X
after 1900, vec off, rebase LEGACY 21749 21763 12 4.6 217.5 1.0X
after 1900, vec off, rebase CORRECTED 22065 22265 188 4.5 220.7 1.0X
after 1900, vec on, rebase EXCEPTION 6613 6766 218 15.1 66.1 3.2X
after 1900, vec on, rebase LEGACY 6134 6243 143 16.3 61.3 3.5X
after 1900, vec on, rebase CORRECTED 6332 6368 54 15.8 63.3 3.4X
before 1900, vec off, rebase LEGACY 24196 24829 579 4.1 242.0 0.9X
before 1900, vec off, rebase CORRECTED 21725 22006 286 4.6 217.3 1.0X
before 1900, vec on, rebase LEGACY 9491 9774 250 10.5 94.9 2.3X
before 1900, vec on, rebase CORRECTED 6136 6229 93 16.3 61.4 3.5X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Save TIMESTAMP_MILLIS to parquet: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1900, noop 3755 3755 0 26.6 37.5 1.0X
before 1900, noop 3619 3619 0 27.6 36.2 1.0X
after 1900, rebase EXCEPTION 21243 21243 0 4.7 212.4 0.2X
after 1900, rebase LEGACY 21808 21808 0 4.6 218.1 0.2X
after 1900, rebase CORRECTED 22194 22194 0 4.5 221.9 0.2X
before 1900, rebase LEGACY 26637 26637 0 3.8 266.4 0.1X
before 1900, rebase CORRECTED 22130 22130 0 4.5 221.3 0.2X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Load TIMESTAMP_MILLIS from parquet: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1900, vec off, rebase EXCEPTION 21977 22216 207 4.6 219.8 1.0X
after 1900, vec off, rebase LEGACY 21929 22352 370 4.6 219.3 1.0X
after 1900, vec off, rebase CORRECTED 22644 22788 171 4.4 226.4 1.0X
after 1900, vec on, rebase EXCEPTION 7719 7818 101 13.0 77.2 2.8X
after 1900, vec on, rebase LEGACY 7736 7800 88 12.9 77.4 2.8X
after 1900, vec on, rebase CORRECTED 7896 8002 107 12.7 79.0 2.8X
before 1900, vec off, rebase LEGACY 25595 25846 218 3.9 255.9 0.9X
before 1900, vec off, rebase CORRECTED 22795 23167 550 4.4 227.9 1.0X
before 1900, vec on, rebase LEGACY 10800 10859 60 9.3 108.0 2.0X
before 1900, vec on, rebase CORRECTED 7546 7635 82 13.3 75.5 2.9X
================================================================================================
Rebasing dates/timestamps in ORC datasource
================================================================================================
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Save DATE to ORC: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1582, noop 25059 25059 0 4.0 250.6 1.0X
before 1582, noop 12893 12893 0 7.8 128.9 1.9X
after 1582 32469 32469 0 3.1 324.7 0.8X
before 1582 25104 25104 0 4.0 251.0 1.0X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Load DATE from ORC: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1582, vec off 11382 11543 194 8.8 113.8 1.0X
after 1582, vec on 4964 5042 99 20.1 49.6 2.3X
before 1582, vec off 12443 12478 31 8.0 124.4 0.9X
before 1582, vec on 5685 5724 34 17.6 56.9 2.0X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Save TIMESTAMP to ORC: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1900, noop 3528 3528 0 28.3 35.3 1.0X
before 1900, noop 3536 3536 0 28.3 35.4 1.0X
after 1900 19807 19807 0 5.0 198.1 0.2X
before 1900 28731 28731 0 3.5 287.3 0.1X
OpenJDK 64-Bit Server VM 11.0.10+9-LTS on Linux 5.4.0-1043-azure
Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Load TIMESTAMP from ORC: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
after 1900, vec off 12936 13032 85 7.7 129.4 1.0X
after 1900, vec on 5957 6006 50 16.8 59.6 2.2X
before 1900, vec off 16065 16193 115 6.2 160.7 0.8X
before 1900, vec on 8611 8836 196 11.6 86.1 1.5X
Reference in a new issue View git blame Copy permalink