elek commented on issue #578: HDDS-3053. Decrease the number of the chunk 
writer threads
URL: https://github.com/apache/hadoop-ozone/pull/578#issuecomment-603816449
 
 
   > Can we make it a data driven decision? I looked at the linked document. I 
didn't understand the test methodology. E.g. when you say 60 thread / spinning 
disk, is that 60 chunk executor threads or 60 client threads?
   
   There was no client thread in this test. The command was mentioned later in 
the doc:
   
   ```
   60 thread / spinning disk
   
   Command:
   
   ozone freon --verbose cmdw -n1000000 -s 1024 -t60
   ```
   
   This is using the ChunkManagerImpl fro 60 threads.
   
   If you are interested about an e2e test, please check the teragen results 
(added in a comment).
   
   > Can we make it a data driven decision?
   
   100% agree. If you need any more date, just let me know. I think we should 
have 60 threads only if we have data as more threads mean more overhead. But if 
we have data to keep it, I am fine with it.
   
   Based on the existing test, 10 threads should be enough as a default (IMHO)

----------------------------------------------------------------
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:
us...@infra.apache.org


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: ozone-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: ozone-issues-h...@hadoop.apache.org

Reply via email to