-
Notifications
You must be signed in to change notification settings - Fork 29k
[SPARK-32227] Fix regression bug in load-spark-env.cmd with Spark 3.0.0 #29044
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
|
cc @srowen |
|
@warrenzhu25 . Could you elaborate about what is the regression you are mentioning here?
|
|
ok to test |
|
@warrenzhu25 it should be best to show the input like your command you tried and output from the command explicitly before/after the fix in PR description especially given that not so many dev people here arguably don't have Windows env. |
|
Test build #125441 has finished for PR 29044 at commit
|
|
@warrenzhu25 As you see, we have a |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Currently, master branch is healthy on Windows.
To be considered as a valid patch, this PR should pass AppVeyor test job on Windows.
As I see, the failure seems unrelated with this change. It seems it couldn't find correct version of hadoop.dll. |
|
It's directly relevant to this PR because your patch is changing
|
|
Please remove |
winutils only impacted by PATH and HADOOP_HOME, and I don't touch both. Also, my change is just reverting into the version as 2.4.4. Could you help rerun the tests? |
|
@dongjoon-hyun Could you help retest this as failing tests might be unrelated? |
|
@warrenzhu25 you can push empty commit to retrigger the tests, or rebase to sync with the master branch. |
|
+1 for @HyukjinKwon 's comment. |
|
Okay, I am debugging a related issue and happened to take a look again for this fix. this fix looks good. AppVeyor currently fails for another issue which I'll probably be able to fix soon. |
|
Test build #126617 has finished for PR 29044 at commit
|
|
Test build #126659 has finished for PR 29044 at commit
|
|
Test build #126670 has finished for PR 29044 at commit
|
|
Test build #126680 has finished for PR 29044 at commit
|
|
@dongjoon-hyun Could you help merge this? |
|
Merged to master and branch-3.0. |
|
Thanks for working on this @warrenzhu25. |
Fix regression bug in load-spark-env.cmd with Spark 3.0.0 cmd doesn't support set env twice. So set `SPARK_ENV_CMD=%SPARK_CONF_DIR%\%SPARK_ENV_CMD%` doesn't take effect, which caused regression. No Manually tested. 1. Create a spark-env.cmd under conf folder. Inside this, `echo spark-env.cmd` 2. Run old load-spark-env.cmd, nothing printed in the output 2. Run fixed load-spark-env.cmd, `spark-env.cmd` showed in the output. Closes #29044 from warrenzhu25/32227. Lead-authored-by: Warren Zhu <[email protected]> Co-authored-by: Warren Zhu <[email protected]> Signed-off-by: HyukjinKwon <[email protected]> (cherry picked from commit 7437720) Signed-off-by: HyukjinKwon <[email protected]>
|
Thank you all. I see that this passed the GitHub Action at 4589e1b . |
What changes were proposed in this pull request?
Fix regression bug in load-spark-env.cmd with Spark 3.0.0
Why are the changes needed?
cmd doesn't support set env twice. So set
SPARK_ENV_CMD=%SPARK_CONF_DIR%\%SPARK_ENV_CMD%doesn't take effect, which caused regression.Does this PR introduce any user-facing change?
No
How was this patch tested?
Manually tested.
echo spark-env.cmdspark-env.cmdshowed in the output.