Zachary Steudel

I want to note that I came across this as well on spark-3.2.1-bin-hadoop3.2 and @fatim's solution worked for me as well. I was on OpenJDK17 and downgraded to OpenJDK11 and...

I enabled github actions on my forked repo but it's not clear to me how to re-kick the failed build now that I've done that. Perhaps a maintainer will know...

I also wanted to note someone attempted to fix this issue with this PR in 2019: https://github.com/apache/spark/pull/25076 and this did get merged, but does not appear to fix the issue....

Wanted to chime in again and note that my team has been using this patched version of the spark-class2.cmd to launch our Spark instances for a couple weeks now and...

Hello all, I'm going to do one final edit on this PR. We found that `set LAUNCHER_OUTPUT=%temp%\spark-class-launcher-output-%RANDOM%%TIME::=0%.txt` still resulted in the occasional race condition error, as we launch tens of...

Alright, that should be my final edit. Like I said, we've had no issues since applying this fix a few months ago and we have since increased the amount of...

Bumping this PR - any other concerns?

Thanks folks, sorry for the delay, got sidetracked and kept forgetting to check this issue again. As mentioned, my company has been using this exact code in our production process...

+1 on this issue. I think usage of `/etc/timezone` should be removed entirely, although I don't have enough knowledge of the history of Linux or potential compatibility issues to be...