[GitHub] [arrow] jorgecarleitao edited a comment on pull request #8635: ARROW-10551: [Rust] Fix unreproducible benches

[GitBox <gi...@apache.org>]

jorgecarleitao edited a comment on pull request #8635:
URL: https://github.com/apache/arrow/pull/8635#issuecomment-725414955


   > Current, code is reseeding after 32 MB of data, and every thread has different randomness. So no data is same as another data and totally different in different use cases.
   
   What I am trying to understand is why this is a concern: in the benchmarks, the data is generated only once, during the setup and not on every iteration of the benchmark. Thus, any randomness will be frozen for every iteration of the benchmark.
   
   If the argument is that two runs of the same benchmark do not agree due to randomness, then I do not understand how having the same seed per thread or different seeds per thread changes this behavior: the seed will still be set when the process starts and will be different per run.
   
   EDIT: @alamb beat me to seconds (and evidence)


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org