Hey @dwagner,
as @jousefm mentioned the error that you experience is caused by a lack of memory on the computing instance. The warnings that you can see in the event log are just additional information as the mesh might be influenced in a negative way by those defects in the geometry, but it does not need to necessarily fail.
In your case you can see at the end of the meshing log that right before the meshing failed the meshing algorithm reported the following amount of cells per refinement level:
After balancing surface refinement iteration 3 : cells:23931933 faces:78383771 points:30768768
Cells per refinement level:
0 189888
1 8723
2 32235
3 127338
4 599305
5 2399851
6 12507644
7 3595245
8 4471704
Surface refinement iteration 4
which adds up to 23931933 total cells. This is a very high number. You can of course now use a 32 core instance instead (the available memory will be approximately doubled), but keep in mind that the computational costs of your simulation will scale with the mesh size. So you might want to adapt your mesh settings if possible. It is always a trade off.
Hope that helps!
Best Alex