Improve balancing of CI test jobs

Adding some related issue in this 'test work allocation optimization' space that have similar intent, ftr and later ease of reference.

I think this is good, probably we just need to try it and see its performance. The test coverage is also good to ensure isolation between the bins.

The only thing I would suggest here is to add a 'weight' column to the debug-discovery report, near to the 'count' column, so to make more visually clear the logic for the sorting, as this is no longer obvious.

Test execution sequence. Tests marked *** will be executed in this PARALLEL BIN #2.
-------------------------------------------------------------------------------------
    Sort  Bin                                 
Bin  Seq  Seq Slow? Group            Cnt Class
-------------------------------------------------------------------------------------
       1      #slow media             52 Drupal\Tests\media\Kernel\MediaAccessControlHandlerTest     
***    2    1 #slow package_manager   30 Drupal\Tests\package_manager\Kernel\SandboxManagerBaseTest  
       3      #slow Database          25 Drupal\Tests\mysql\Kernel\mysql\SchemaTest                  
       4      #slow Entity            22 …elTests\Core\Entity\EntityDefinitionUpdateMultipleTypesTest

Tweaking the weight of the #slow tests would probably make sense if we end up with major WT differences between bins of the same test type.

Let’s see first; time will tell. RTBC

I have considered opening an issue to add ##slow and ###slow multiple times but assumed that people would point and laugh at me for it, this wouldn't necessarily stop me from doing it.

This looks great to me, good way to get both walltimes and CPU usage down at the same time without needing to refactor any tests at all!

We could also potentially look at boosting the weight for lower method-count slow tests a bit - those tend to have the slowest individual methods. e.g. assuming average method times of 7s, a test with one 20s method doesn't need #slow (although it's annoying if it's the last test to run, but usually not a big deal), a test with ten 20s methods generally does, a test with one 120s method definitely does. Might open a follow-up to experiment a bit, but easier to do that with this already in.

#3615585: Optimize EntityDefinitionUpdateMultipleTypesTest improves EntityDefinitionUpdateMultipleTypesTest performance and needs another review.

Committed/pushed to main and 12.0.x, thanks!

Opened #3632681: Tweak test weights for #slow issues for the idea in #12.

#14: also #3504223: Introduce a “weight“ to #slow tests was meaning that.