ahshahid opened a new pull request, #53100:
URL: https://github.com/apache/spark/pull/53100

   …ed by buffers of the spawned process, not being consumed.
   
   ### What changes were proposed in this pull request?
   Added code in the client VM to consume the output and error stream of the 
connect server process spawned by it, so that spawned process's buffers are 
cleared..
   
   
   ### Why are the changes needed?
   The Connect Server process spawned by the test using the class 
RemoteSparkSession, by default ( non debug mode) , does not consume the spawned 
process's outout and error buffers. In certain situations ( extra logging ) it 
can cause deadlock, as the synchronous  thread doing the logging acquires the 
lock in the Log4J's class, but is not able to write as the buffers are full... 
which causes other threads in the server vm to block , waiting to acquire the 
lock.
   The stacktraces of the deadlock situation are available in the corresponding 
ticket.
   
   
   ### Does this PR introduce _any_ user-facing change?
   No
   
   
   ### How was this patch tested?
   All the existing tests ( including the JavaEncoderSuite) passes with the 
code of cleaning up of the buffers on my local checxkout.
   
   
   ### Was this patch authored or co-authored using generative AI tooling?
   No


-- 
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