Github user bersprockets commented on the issue:
https://github.com/apache/spark/pull/20519
>corner case described in the JIRA
I don't have a good feeling for how often people use python site
customizations. I did encounter a real life case that caused the failure
described in the Jira.
This is your area, I will leave it up to you. I still have the first PR
where we just better report this case rather than try to fix it.
The logic of the code is largely the same:
- launch the daemon
- read the port number from a DataInputStream connected to the daemon.
The steps to set up a socket, and the try/catches required, add quite a few
lines, even though the end result is the same (read an int from a
DataInputStream connected to the daemon).
That being said, the code does add some complexities and timing differences.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
