vtutrinov opened a new pull request, #5398:
URL: https://github.com/apache/ozone/pull/5398

   ## What changes were proposed in this pull request?
   
   The OM's org.apache.hadoop.ozone.om.OzoneManagerServiceGrpc#submitRequest 
method uses UUID.randomUUID() (through 
org.apache.hadoop.ipc.ClientId#getClientId) to set the clientId for an 
omRequest. The latter has a synchronized block inside that leads to high lock 
contention in case of a big amount of tiny files (1K) read requests (we were 
testing a read of 15mln 1KiB files from 3 clients in 50 threads, 150 threads in 
total). Hence, the clientId generator should be thread-local
   
   ## What is the link to the Apache JIRA
   
   https://issues.apache.org/jira/browse/HDDS-9379
   
   ## How was this patch tested?
   
   Manual test of reading of 15mln 1KiB files
   


-- 
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