|Jul 25, 2014|
|5 years 3 months ||SnowpackInterface: It seems that it's enough to call the
runoff->output(Date&) method from the master node only, since
data is supplied solely through runoff->fillHydroTables.
|5 years 3 months ||SnowpackInterface: now runs with MPI and scales very well - however
two issues remain: interdependecy of modules (runoff is called from
within snowpack) and the writing of data, that is distributed on
multiple MPI nodes (writeOutputSNO, writing runof
|Jul 24, 2014|
|5 years 3 months ||SnowpackInterface: Cleanup before inducing MPI
|5 years 3 months ||MeteoObj: Sanitizing for MPI usage
|5 years 3 months ||MeteoObj: There was a misconception as to which date has been
prepared. The non-master MPI nodes simply assume that data is
available for the correct date. Alternatively we could broadcast the
date available in an async fashion.
|5 years 3 months ||MPIControl: Adding a barrier method that calls MPI_Barrier() or does
nothing in case MPI is not active. This can be used to synchronize
anywhere in the code, mostly useful for debugging.
|5 years 3 months ||Cosmetics, reducing output from ViewFactors
|5 years 3 months ||Reverting changes made to ViewFactors class. There was an error when
computing the view factors - further investigation required.
|5 years 3 months ||Adding CMake scritps for finding PETSc. In any case the environment
variables PETSC_DIR and PETSC_ARCH need to be set!
|5 years 3 months ||main / ebalance: cosmetics
|Jul 23, 2014|
|5 years 3 months ||AlpineControl: Changing log messages to adhere to [i],[w],[e]
|5 years 3 months ||TerrainRadiationPETSc: Getting rid of too verbose output and
cleaning up code.
IMPORTANT: onyl version PETSc 3.4.5 and earlier will work seamlessly
with this code.
|5 years 3 months ||TerrainRadiation: Adding the TerrainRadiation method on basis of
PETSc encapsuled in a class called TerrainRadiationPETSc. To
activate this option, which produces the same out put as the default
TerrainRadiation method, but is more efficient because
|Jul 21, 2014|
|5 years 3 months ||Getting rid of some typedefs for longuint, keeping it in RunOff,
where it actually makes sense.
|5 years 3 months ||MPIControl: getting rid of a warning
|5 years 3 months ||ViewFactors: Trying to introduce size_t in a stringent fashion.
|5 years 3 months ||EnergyBalance: Bugfix, the indices for the OpemMP threads were not
mapped onto the correct DEM portion.
|5 years 3 months ||MPIControl: Adding OpenMP methods int thread() and bool openmp(),
which return the current thread number or zero and whether OpenMP
support has been activated during compilation or not.
|5 years 3 months ||SnowpackInterface: no ifdefs for openmp, everything is handled in
|5 years 3 months ||EnergyBalance: Reintroducing OpenMP - now both OpenMP and MPI can
coexist in one binary peacefully. Every MPI process gets a certain
slice of the DEM and divides that DEM locally according to the
number of workers configured.