Important fix of LOCAL_ORDER @jerqi

Codecov Report

Merging #319 (117cc82) into master (4a3d2be) will increase coverage by 0.07%.
The diff coverage is 100.00%.

@@             Coverage Diff              @@
##             master     #319      +/-   ##
============================================
+ Coverage     61.01%   61.09%   +0.07%     
- Complexity     1489     1494       +5     
============================================
  Files           185      185              
  Lines          9314     9325      +11     
  Branches        900      903       +3     
============================================
+ Hits           5683     5697      +14     
+ Misses         3326     3323       -3     
  Partials        305      305              

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

Error: Failed to execute goal on project uniffle-parent: Could not resolve dependencies for project org.apache.uniffle:uniffle-parent:pom:0.7.0-snapshot: Failed to collect dependencies at org.slf4j:slf4j-log4j12:jar:1.7.25: Failed to read artifact descriptor for org.slf4j:slf4j-log4j12:jar:1.7.25: Could not transfer artifact org.slf4j:slf4j-log4j12:pom:1.7.25 from/to central (https://repo1.maven.org/maven2): transfer failed for https://repo1.maven.org/maven2/org/slf4j/slf4j-log4j12/1.7.25/slf4j-log4j12-1.7.25.pom: Connect to repo1.maven.org:443 [repo1.maven.org/146.75.28.209] failed: Connection timed out (Connection timed out) -> [Help 1]

We should add a ut to cover this case, because our exist tests can't find this problem.

Emm... Actually AQESkewedJoinWithLocalOrderTest could be cover the bug of expectedTaskIds .
However because the distribution type of LOCAL_ORDER is invalid with the incorrect initialization of shuffleReadClientImpl, it fall back to the NORMAL segment splitter, which cause the AQESkewedJoinWithLocalOrderTest run successfully.

it's a coincidence

Emm... Actually AQESkewedJoinWithLocalOrderTest could be cover the bug of expectedTaskIds . However because the distribution type of LOCAL_ORDER is invalid with the incorrect initialization of shuffleReadClientImpl, it fall back to the NORMAL segment splitter, which cause the AQESkewedJoinWithLocalOrderTest run successfully.

it's a coincidence

OK. It don't influence our performance test result, do it?

Emm... Actually AQESkewedJoinWithLocalOrderTest could be cover the bug of expectedTaskIds . However because the distribution type of LOCAL_ORDER is invalid with the incorrect initialization of shuffleReadClientImpl, it fall back to the NORMAL segment splitter, which cause the AQESkewedJoinWithLocalOrderTest run successfully.
it's a coincidence

OK. It don't influence our performance test result, do it?

Emm... All bugs are caused by #137 subsequent refactors of changing the [startMapId, endMapid) to taskIds bitmap.

The performance results are based on the initial commit. So the result is OK.

Emm... Actually AQESkewedJoinWithLocalOrderTest could be cover the bug of expectedTaskIds . However because the distribution type of LOCAL_ORDER is invalid with the incorrect initialization of shuffleReadClientImpl, it fall back to the NORMAL segment splitter, which cause the AQESkewedJoinWithLocalOrderTest run successfully.
it's a coincidence

OK. It don't influence our performance test result, do it?

Emm... All bugs are caused by #137 subsequent refactors of changing the [startMapId, endMapid) to taskIds bitmap.

The performance results are based on the initial commit. So the result is OK.

OK. LGTM, thanks @zuston