Re: [deal.II] Distributing objects on cluster nodes according to distributed triangulation (MPI)

2019-09-04 Thread Konrad
> > boost::mpi is a nice idea in principle, but it's not particularly well > designed and few people use it. Some reasons are listed here: > > > https://scicomp.stackexchange.com/questions/3019/boostmpi-or-c-mpi-for-high-performance-scientific-applications > > > Best > W. > I think it

Re: [deal.II] Distributing objects on cluster nodes according to distributed triangulation (MPI)

2019-09-03 Thread Wolfgang Bangerth
On 9/3/19 2:40 AM, Konrad wrote: > > I have a little technical question though: Can I attach the mpi_communicators > that you use in the deal.ii tutorials also somehow to a > boost::mpi::communicator? > > https://www.boost.org/doc/libs/1_48_0/doc/html/boost/mpi/communicator.html > > Never

Re: [deal.II] Distributing objects on cluster nodes according to distributed triangulation (MPI)

2019-09-03 Thread Konrad
Thank you, Wolfgang, problem solved. But in a much simpler way. My basis object is not easily serializable so I have to recreate it and destroy the old one anyway (for now) once cells are moved to other nodes. So I do not send send them but rather create them and work with them on each node (if

Re: [deal.II] Distributing objects on cluster nodes according to distributed triangulation (MPI)

2019-08-23 Thread Wolfgang Bangerth
On 8/23/19 7:33 AM, Konrad wrote: > > Thanks, I see this class makes it possible to identify cells globally in a > distributed triangulation. I don't see though how the objects in the std::map > are distributed among cluster nodes as the cells are. Am I missing something > here? Yes, the data

Re: [deal.II] Distributing objects on cluster nodes according to distributed triangulation (MPI)

2019-08-23 Thread Konrad
Hi Daniel, > You could create an std::map using CellId ( > https://www.dealii.org/developer/doxygen/deal.II/classCellId.html) as key > if you need to identify cells uniquely globally. > > Best, > Daniel > Thanks, I see this class makes it possible to identify cells globally in a

Re: [deal.II] Distributing objects on cluster nodes according to distributed triangulation (MPI)

2019-08-22 Thread Daniel Arndt
Konrad, I have a little conceptual question. Maybe it is dumb but I am new to MPI... > > Say I have a triangulation of type > *parallel::distributed::Triangulation* that gets distributed over N > nodes on a cluster. Each active cell should own (or point to) an object > that makes up a (modified

[deal.II] Distributing objects on cluster nodes according to distributed triangulation (MPI)

2019-08-22 Thread Konrad
Hi deal.ii community, I have a little conceptual question. Maybe it is dumb but I am new to MPI... Say I have a triangulation of type *parallel::distributed::Triangulation* that gets distributed over N nodes on a cluster. Each active cell should own (or point to) an object that makes up a