GH-43502: [Java] Fix Java JNI / AMD64 manylinux2014 Java JNI test not test dataset module

[vibhatha]

Rationale for this change

JNI tests had a typo in the script where we have included gandiva instead of dataset.

What changes are included in this PR?

This PR fixes that typo in the current version.

Are these changes tested?

Yes, by existing tests and CIs.

Are there any user-facing changes?

No

• GitHub Issue: [Java] Fix Java JNI / AMD64 manylinux2014 Java JNI test not test dataset module #43502

[vibhatha]

@github-actions crossbow submit -g java

[github-actions]

⚠️ GitHub issue #43502 has been automatically assigned in GitHub to PR creator.

[github-actions]

Revision: a394350

Submitted crossbow builds: ursacomputing/crossbow @ actions-a30ce59033

Task	Status
java-jars
test-conda-python-3.11-spark-master
verify-rc-source-java-linux-almalinux-8-amd64
verify-rc-source-java-linux-conda-latest-amd64
verify-rc-source-java-linux-ubuntu-20.04-amd64
verify-rc-source-java-linux-ubuntu-22.04-amd64
verify-rc-source-java-macos-amd64

[jinchengchenghh]

And it still has a problem , https://github.com/apache/arrow/blob/main/ci/docker/java-jni-manylinux-201x.dockerfile#L53, ARROW_JAVA_JNI set, so jni build exists, but jni test detect ARROW_DATASET

[vibhatha]

And it still has a problem , https://github.com/apache/arrow/blob/main/ci/docker/java-jni-manylinux-201x.dockerfile#L53, ARROW_JAVA_JNI set, so jni build exists, but jni test detect ARROW_DATASET

I didn't get it clearly, could you please elaborate a bit?

[jinchengchenghh]

It check the ARROW_DATASET, but it is OFF now, so as ARROW_GANDIVA https://github.com/apache/arrow/blob/main/ci/scripts/java_test.sh#L41

[vibhatha]

It check the ARROW_DATASET, but it is OFF now, so as ARROW_GANDIVA https://github.com/apache/arrow/blob/main/ci/scripts/java_test.sh#L41

Let me check, though if this is true, we have had the CIs running without any of these checks 🤔

[vibhatha]

@jinchengchenghh on top of that there is already a CI failure for dataset.

[jinchengchenghh]

It check the ARROW_DATASET, but it is OFF now, so as ARROW_GANDIVA https://github.com/apache/arrow/blob/main/ci/scripts/java_test.sh#L41

Let me check, though if this is true, we have had the CIs running without any of these checks 🤔

I think you are right, not sure where to set the environment, but the test really triggers.

I will try to fix the exception, my IDE sync cause some problems

[jinchengchenghh]

Can you mark it as ignore? I will draft a new PR to solve it.

[vibhatha]

It check the ARROW_DATASET, but it is OFF now, so as ARROW_GANDIVA https://github.com/apache/arrow/blob/main/ci/scripts/java_test.sh#L41

Let me check, though if this is true, we have had the CIs running without any of these checks 🤔

I think you are right, not sure where to set the environment, but the test really triggers.

I will take a look at that.

I will try to fix the exception, my IDE sync cause some problems

Thanks, that would be wonderful, but I am not sure if this is related to the recent PRs (didn't check yet).

[vibhatha]

Can you mark it as ignore? I will draft a new PR to solve it.

would you like to push the patch to this branch?

[jinchengchenghh]

#43506

[vibhatha]

Btw, does that failing test in the CI work locally?

[jinchengchenghh]

Do you know how to push to this branch? I don't think I have write access to vibhatha:gh-43502

[jinchengchenghh]

#41646 skip tests java-jars, maybe we still need environment ARROW_DATASET to trigger it in Java JNI / AMD64 manylinux2014 Java JNI test

[vibhatha]

Do you know how to push to this branch? I don't think I have write access to vibhatha:gh-43502

I will add the rights.

[vibhatha]

@jinchengchenghh I have added you and you should have commit rights, please check.

[jinchengchenghh]

ok thanks! I will draft a PR as soon as possible

[jinchengchenghh]

I investigate the question again, the exception should not occur, looks like the linked library is not right.

TestFragmentScanOptions.testCsvConvertOptions:80 expected: <[Id: Int(32, true), Name: Utf8, Language: Utf8]> but was: <[Id;Name;Language: Utf8]>

Can anyone help to check it？

[vibhatha]

I investigate the question again, the exception should not occur, looks like the linked library is not right.

TestFragmentScanOptions.testCsvConvertOptions:80 expected: <[Id: Int(32, true), Name: Utf8, Language: Utf8]> but was: <[Id;Name;Language: Utf8]>

Can anyone help to check it？

let me check this locally.

[vibhatha]

@github-actions crossbow submit java-jars

[vibhatha]

@github-actions crossbow submit -g java

[github-actions]

Revision: cb36191

Submitted crossbow builds: ursacomputing/crossbow @ actions-92c9a56dee

Task	Status
java-jars
test-conda-python-3.11-spark-master
verify-rc-source-java-linux-almalinux-8-amd64
verify-rc-source-java-linux-conda-latest-amd64
verify-rc-source-java-linux-ubuntu-20.04-amd64
verify-rc-source-java-linux-ubuntu-22.04-amd64
verify-rc-source-java-macos-amd64

[vibhatha]

@kou I updated as you suggested, also created an issue to track and update once the LLVM issue is resolved.

[danepitkin]

@github-actions crossbow submit java-jars

[github-actions]

Revision: cb36191

Submitted crossbow builds: ursacomputing/crossbow @ actions-67f1c3bfef

Task	Status
java-jars

[vibhatha]

@danepitkin shall we re-run the java-jars? I am not sure what that failure is.

[kou]

Hmm. self-hosted arm64 Linux runner may have a problem:

arrow/dev/tasks/java-jars/github.yml

Line 43 in 2e1b245

- runs_on: ["self-hosted", "Linux", "arm64"]

Disk size?

[vibhatha]

@kou honestly not sure...

[danepitkin]

I think the previous java-jars successful run proves its fine. However the Java JNI CI job is failing. Did we want to fix the TestCsvFragment options in this PR?

[vibhatha]

I think the previous java-jars successful run proves its fine. However the Java JNI CI job is failing. Did we want to fix the TestCsvFragment options in this PR?

Wasn't the objective in this PR though.

[danepitkin]

@github-actions crossbow submit java-jars

[github-actions]

Revision: cb36191

Submitted crossbow builds: ursacomputing/crossbow @ actions-f25c638877

Task	Status
java-jars

[danepitkin]

Gandiva builds, java-jars now failing on #43506

[vibhatha]

Thanks @danepitkin

[conbench-apache-arrow]

After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit cdfdfb8.

There were no benchmark performance regressions. 🎉

The full Conbench report has more details. It also includes information about 3 possible false positives for unstable benchmarks that are known to sometimes produce them.