[AURON #1595] Enforce skiptest=false to not miss failing UT

[ShreyeshArangath]

Which issue does this PR close?

Closes #1595

Rationale for this change

• Ensure UTs run in pre-builds to catch regressions before merging.
• Correct Exchange conversion behavior to respect feature flags/shuffle manager.

What changes are included in this PR?

• CI: Run tests in pre-builds: append --skiptests false
• Core: Fix failing UTs

Are there any user-facing changes?

• Behavior corrected in native Exchange conversion which was causing test failures
• No public API changes.

How was this patch tested?

• CI pre-build runs UTs via --skiptests false.

Note: I'll create another PR to add support for publishing test reports (draft: PR)

[ShreyeshArangath]

The Celeborn TPC-DS failure seems unrelated to this PR (OOM), we might have to investigate why this is happening. If anyone has a lead on this, please do let me know.

Caused by: java.lang.RuntimeException: poll record batch error: Execution error: native execution panics: Execution error: Execution error: output_with_sender[Shuffle] error: Execution error: output_with_sender[Agg] error: Execution error: output_with_sender[Expand] error: Execution error: output_with_sender[Expand]: output() returns error: Execution error: Execution error: output_with_sender[Project] error: Execution error: output_with_sender[Agg] error: Execution error: output_with_sender[IpcReader] error: Execution error: output_with_sender[IpcReader]: output() returns error: External error: Java exception thrown at native-engine/datafusion-ext-plans/src/ipc_reader_exec.rs:216: java.io.IOException: External error: Java exception thrown at native-engine/datafusion-ext-plans/src/ipc_reader_exec.rs:339: java.lang.OutOfMemoryError: Java heap space
at org.apache.auron.jni.JniBridge.nextBatch(Native Method)
at org.apache.auron.jni.AuronCallNativeWrapper.loadNextBatch(AuronCallNativeWrapper.java:118)
... 32 more

Edit: looks like the re-run succeeded; we should still look into tuning the TPC-DS execution. Maybe, we are running very close to the heap limits.

[slfan1989]

LGTM

[guixiaowen]

LGTM

[ShreyeshArangath]

@guixiaowen can you help fix the failing UT introduced by #1585? Thanks

[ShreyeshArangath]

Updated the existing tests to not fail

[ShreyeshArangath]

@SteNicholas @merrily01 Can you please help me merge this to prevent any further unit test failures?

[SteNicholas]

cc @richox.

[ShreyeshArangath]

@richox / @cxzl25, could you please take a look?