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 @ 2.60GHz
   Benchmark KryoPool vs old"pool of 1" implementation:  Best Time(ms)   Avg 
Time(ms)   Stdev(ms)    Rate(M/s)   Per Row(ns)   Relative
   
-----------------------------------------------------------------------------------------------------------------------------------
   KryoPool:true                                                 9375          
12171         NaN          0.0    18750400.9       1.0X
   KryoPool:false                                               13849          
16799         NaN          0.0    27697646.0       0.7X
   ```
   
   Based on this it looks like the branch is branch might is a bit faster. But 
I think it might also be in noise territory and that one would need a more 
specific benchmark that creates a lot of small streams for the difference to 
show up. I think it only expected to be order of percent better in the "worst 
case" when we are creating lots of small streams.


-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to