Skip to content
New issue

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

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

Already on GitHub? Sign in to your account

[GLUTEN-6887][VL] Daily Update Velox Version (2024_08_19) #6910

Merged
merged 1 commit into from
Aug 19, 2024

Conversation

GlutenPerfBot
Copy link
Contributor

Upstream Velox's New Commits:

54b16e041 by Ke Wang, Refactor WriterOptions processConfigs function (10762)
2b7f559a4 by Pedro Eugenio Rocha Pedreira, Fix std::shared_mutex missing includes (10772)
6b9806069 by xiaoxmeng, Fix flaky FileInputStreamTest.stats (10771)
7fd6273e5 by Chengcheng Jin, Move decimal-to-string API to DecimalUtil (10477)
d23dde5aa by Wei He, Avoid generating unsupported data types for JoinFuzzer with PQR (10662)
d805d31a0 by Jialiang Tan, Apply grow target adjustment to reduce arbitration count (10170)
e7ffae924 by xiaoxmeng, Remove byte input stream legacy code (10761)
44fb107d0 by Wei He, Use ASSERT_STREQ when comparing C-strings (10759)
c701f8308 by Satadru Pan, Match Velox hmac_sha/md5 functions behavior with Java on empty keys (10744)
d08d69469 by xiaoxmeng, Some window operator code cleanup (10764)
bc274f2a8 by Bikramjeet Vig, Ensure flattening an array or map does not result in an oversized values vector (10760)

@zhouyuan
Copy link
Contributor

@GlutenPerfBot benchmark

@GlutenPerfBot
Copy link
Contributor Author

ACK, will benchmark TPCH/DS on this pull request

@zhouyuan
Copy link
Contributor

#6887

@GlutenPerfBot
Copy link
Contributor Author

===== Performance report for TPCH SF2000 with Velox backend, for reference only ====

query log/native_6910_time.csv log/native_master_08_18_2024_9fcd4885c7_time.csv difference percentage
q1 39.42 40.01 0.589 101.50%
q2 30.56 28.03 -2.535 91.71%
q3 53.89 53.67 -0.219 99.59%
q4 40.88 42.78 1.897 104.64%
q5 108.87 104.21 -4.666 95.71%
q6 11.60 11.23 -0.365 96.85%
q7 117.15 115.52 -1.630 98.61%
q8 114.69 116.72 2.024 101.76%
q9 169.01 168.24 -0.771 99.54%
q10 64.89 66.68 1.795 102.77%
q11 27.76 30.17 2.405 108.66%
q12 33.00 31.20 -1.796 94.56%
q13 51.40 51.65 0.253 100.49%
q14 24.35 26.30 1.947 107.99%
q15 55.54 54.84 -0.700 98.74%
q16 19.87 19.79 -0.081 99.59%
q17 133.11 131.24 -1.867 98.60%
q18 197.98 197.27 -0.706 99.64%
q19 26.20 25.90 -0.301 98.85%
q20 41.59 40.46 -1.127 97.29%
q21 387.36 386.69 -0.666 99.83%
q22 15.60 15.31 -0.288 98.16%
total 1764.74 1757.93 -6.807 99.61%

@zhouyuan
Copy link
Contributor

note: there's one pending fix not included
facebookincubator/velox#10768

@zhouyuan zhouyuan merged commit bd32c76 into apache:main Aug 19, 2024
40 of 41 checks passed
sharkdtu pushed a commit to sharkdtu/gluten that referenced this pull request Nov 11, 2024
Velox Changes:
```
54b16e041 by Ke Wang, Refactor WriterOptions processConfigs function (10762)
2b7f559a4 by Pedro Eugenio Rocha Pedreira, Fix std::shared_mutex missing includes (10772)
6b9806069 by xiaoxmeng, Fix flaky FileInputStreamTest.stats (10771)
7fd6273e5 by Chengcheng Jin, Move decimal-to-string API to DecimalUtil (10477)
d23dde5aa by Wei He, Avoid generating unsupported data types for JoinFuzzer with PQR (10662)
d805d31a0 by Jialiang Tan, Apply grow target adjustment to reduce arbitration count (10170)
e7ffae924 by xiaoxmeng, Remove byte input stream legacy code (10761)
44fb107d0 by Wei He, Use ASSERT_STREQ when comparing C-strings (10759)
c701f8308 by Satadru Pan, Match Velox hmac_sha/md5 functions behavior with Java on empty keys (10744)
d08d69469 by xiaoxmeng, Some window operator code cleanup (10764)
bc274f2a8 by Bikramjeet Vig, Ensure flattening an array or map does not result in an oversized values vector (10760)
```
Signed-off-by: glutenperfbot <[email protected]>
Co-authored-by: glutenperfbot <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants