Collection files not loaded properly in parallel
This issue was created automatically from an original Mantis Issue. Further discussion may take place here.
If a Collection (pvd file) contains T time steps, and each time step is a multiblock file (vtmb), then when loading in parallel, we would expect that if the multiblock data consists of N blocks and we load on N processors, then one block per processor will be allocated.
In fact, paraview splits each block between all N processors so we have NxN pieces. This impacts performance drastically (IO is terrible) and makes paraview useless for this kind of data.