Dear Wolfgang, excellent! Thank you so much!
Even though I'm not calling GridTools::partition_triangulation() directly, following your hint I found out that the behavior can be controlled directly in the constructor of parallel::shared::Triangulation. So I changed triangulation(mpi_communicator,Triangulation<dim>::maximum_smoothing), to triangulation(mpi_communicator,Triangulation<dim>::maximum_smoothing,false,parallel::shared::Triangulation<dim>::partition_zorder), Best, Alex Wolfgang Bangerth schrieb am Montag, 21. Oktober 2024 um 19:28:13 UTC+2: > > On 10/21/24 03:58, 'Alexander Greiner' via deal.II User Group wrote: > > > > For step-17 I get the same error message indicating that the memory > > allocation failed, for step-18 and my personal code it just fills up the > > system RAM until one aborts it. (Everything works using only 1 core > > though.) Similarly, step-42 for example runs perfectly fine. > > Since the issue seems to be spack miscompiling METIS, I don't really > > expect there is a solution for that, other than trying to recompile > > everything or switching to parallel::distributed::Triangulation. Still, > > I'm very grateful for any suggestions! > > Somewhere in step-18 and in your own code, you are calling > GridTools::partition_triangulation(): > > https://www.dealii.org/current/doxygen/deal.II/namespaceGridTools.html#a99eba8e3b388258eda37a2724579dd1d > It takes a defaulted third argument that you could choose other than > METIS and see whether that addresses the issue by simply not using METIS. > > Best > W. > -- The deal.II project is located at http://www.dealii.org/ For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en --- You received this message because you are subscribed to the Google Groups "deal.II User Group" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/caefaf95-b667-40a5-9852-f905e7e573a8n%40googlegroups.com.
