Github user shivaram commented on the issue:
https://github.com/apache/spark/pull/14741
The value of `what` shouldn't change if we are retrying the read. If it was
`integer` it should remain `integer`. There is a corner case of what happens if
we say read the first byte of the integer and not the remaining 3 bytes -- One
way to handle that would be to read everything into a raw byte stream and do
the cast to integer later. But I think we can postpone this to a later PR.
Regarding your earlier question the easiest way to retry is to run a while
loop around `readBin` while there are more bytes to read. Also we can have a
fixed upper bound and throw an error after those many retries.
Also I dont think SIGKIlLL is the appropriate signal to send - You might
want to send SIGUSR1 (the code is 10 according to
http://linuxandfriends.com/linux-signals/)
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
