[GitHub] [spark] eejbyfeldt commented on pull request #38428: [SPARK-40912][CORE]Overhead of Exceptions in KryoDeserializationStream

2023-05-08 Thread via GitHub
eejbyfeldt commented on PR #38428: URL: https://github.com/apache/spark/pull/38428#issuecomment-1537845261 My plan was to update the benchmarks. But I did not get around to uploading the results until today. But now the branch should be updated with an up to date run. -- This is an

[GitHub] [spark] eejbyfeldt commented on pull request #38428: [SPARK-40912][CORE]Overhead of Exceptions in KryoDeserializationStream

2023-05-03 Thread via GitHub
eejbyfeldt commented on PR #38428: URL: https://github.com/apache/spark/pull/38428#issuecomment-1532703025 > This slipped through my radar, thanks for the ping. Can you update to latest master @eejbyfeldt ? Updated. Will also rerun the benchmarks since those had conflicts. -- This

[GitHub] [spark] eejbyfeldt commented on pull request #38428: [SPARK-40912][CORE]Overhead of Exceptions in KryoDeserializationStream

2023-04-26 Thread via GitHub
eejbyfeldt commented on PR #38428: URL: https://github.com/apache/spark/pull/38428#issuecomment-1522936049 Ping: Anyone has more thoughts or comments on this. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL

[GitHub] [spark] eejbyfeldt commented on pull request #38428: [SPARK-40912][CORE]Overhead of Exceptions in KryoDeserializationStream

2023-01-31 Thread via GitHub
eejbyfeldt commented on PR #38428: URL: https://github.com/apache/spark/pull/38428#issuecomment-1410514876 > Wait, am I reading this wrong, or is the new approach slower? useIterator=true cases seem to take a tiny bit longer. If that's right, what's the benefit? I think you are

[GitHub] [spark] eejbyfeldt commented on pull request #38428: [SPARK-40912][CORE]Overhead of Exceptions in KryoDeserializationStream

2023-01-28 Thread via GitHub
eejbyfeldt commented on PR #38428: URL: https://github.com/apache/spark/pull/38428#issuecomment-1407369412 Rebased and add the results from all the benchmarks. To me is not clear that the regression or we are just looking at noise. Is there some way to make the bechmark more

[GitHub] [spark] eejbyfeldt commented on pull request #38428: [SPARK-40912][CORE]Overhead of Exceptions in KryoDeserializationStream

2023-01-12 Thread GitBox
eejbyfeldt commented on PR #38428: URL: https://github.com/apache/spark/pull/38428#issuecomment-1380312253 > The PR as such looks reasonable to me - can we add a test to explicitly test for EOF behavior ? @mridulm I added a spec for this in:

[GitHub] [spark] eejbyfeldt commented on pull request #38428: [SPARK-40912][CORE]Overhead of Exceptions in KryoDeserializationStream

2023-01-11 Thread GitBox
eejbyfeldt commented on PR #38428: URL: https://github.com/apache/spark/pull/38428#issuecomment-1378799229 I ran the master branch again and used an executor with the same cpu. ``` OpenJDK 64-Bit Server VM 1.8.0_352-b08 on Linux 5.15.0-1023-azure Intel(R) Xeon(R) Platinum 8272CL CPU

[GitHub] [spark] eejbyfeldt commented on pull request #38428: [SPARK-40912][CORE]Overhead of Exceptions in KryoDeserializationStream

2023-01-11 Thread GitBox
eejbyfeldt commented on PR #38428: URL: https://github.com/apache/spark/pull/38428#issuecomment-1378675283 So I ran the benchmark: ``` ``` -- 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