[Core][MPI] Update test suite for DistributedVectorExporter

[loumalouomega]

📝 Description

This PR updates the test suite for the DistributedVectorExporter in the file test_distributed_exporter.cpp. The change modifies the test case's suite from KratosCoreFastSuite to KratosMPICoreFastSuite.

NOTE: The test is failing, like the suite was wrong it was not running in the CI, so we lost track when stopped working, @RiccardoRossi took a look

🆕 Changelog

• Update suite

[loumalouomega]

You can see the false positives @jcotela : https://github.com/KratosMultiphysics/Kratos/actions/runs/5055290302/jobs/9071307435

 |  /           |                  
 ' /   __| _` | __|  _ \   __|    
 . \  |   (   | |   (   |\__ \  
_|\_\_|  \__,_|\__|\___/ ____/
           Multi-Physics 9.3."1"--2eaf8a4-Custom-x86_64
           Compiled for GNU/Linux and Python3.10 with Clang-14.0
Compiled with threading and MPI support.
Maximum number of threads: 1.
MPI world size:         2.
Process Id: 3883 
TestAssignMPIUniqueModelPartCollectionTagUtility                        OK.
TestBenchmarkDistributedGraphConstructionMPI                            OK.
...blah blah
TestDistributedVectorExporter                                           FAILED!
...blah blah
Ran 236 of 1544 test cases in 0.191161s (2 skipped). 1 failed:
    TestDistributedVectorExporter Failed with messages: 
From rank 0:
Test was reported as successful on this rank, but failed on a different rank.
From rank 1:
Error: Check failed because x[x.GetNumbering().LocalId(total_size-1)] = 20 is not near to 15.0 = 15 within the tolerance 1e-14
in kratos/mpi/tests/cpp_tests/sources/test_distributed_exporter.cpp:60: virtual void Testing::TestDistributedVectorExporter::TestFunction()

[loumalouomega]

After the merge with master should break

[loumalouomega]

@RiccardoRossi you can see that with more than 2 partitions the test fails

[loumalouomega]

@RiccardoRossi check this

[loumalouomega]

Reping @RiccardoRossi

[RiccardoRossi]

i have no time to check this now... can u try to debug what happens here?

[loumalouomega]

i have no time to check this now... can u try to debug what happens here?

I already tried and I could find what was the problem. I don't have the time neither for more debug.

[RiccardoRossi]

I took a quick look into this, and my impression is that x is not initialized ... so it addes the 5,9,15 to a vector that may contain anything...

[RiccardoRossi]

@loumalouomega it looks like this fixes it (pending all the tests are finished)

[RiccardoRossi]

@loumalouomega you opened this, so I can approve it. Feel free to merge it when you like