MemoryUsage

Memory usage statistics for the running simulation.

This postprocessor collects various memory usage metrics:

physical memory (most relevant, tied to hardware RAM limitation)
virtual memory
major page faults (how often disk swap is accessed)

The units for memory default to MegaBytes, but users can report usage in other units through the mem_units parameter: (bytes, kilobytes, megabytes, gigabytes).

The data can be reduced in parallel as

maximum out of all MPI ranks
minimum out of all MPI ranks
average over all MPI ranks
total sum over all MPI ranks (default)

Physical memory statistics are available on Mac and Linux, virtual memory and page fault data is Linux only.

This postprocessor can be executed multiple times per timestep and by default aggregates the per-process peak value of the chosen metric, which then can be output on TIMESTEP_END.

note

Until October 2018 this Postprocessor defaulted to reporting virtual memory. This was changed to the more relavant physical memory to avoid misleading benchmark results to be generated.

For a VectorPostprocessor that provides detailed per MPI rank memory statistics see VectorMemoryUsage.

Implementation

The /proc/self/status files is checked first. This file only exists on Linux systems and contains several columns with process specific statistics. On mac systems a conditionally compiled code (#ifdef __APPLE__) block uses a mach kernel API function task_info to obtain the memory sizes of the current process.

Input Parameters

execute_onTIMESTEP_ENDThe list of flag(s) indicating when this object should be executed, the available options include NONE, INITIAL, LINEAR, NONLINEAR, TIMESTEP_END, TIMESTEP_BEGIN, FINAL, CUSTOM.
Default:TIMESTEP_END
C++ Type:ExecFlagEnum
Options:NONE INITIAL LINEAR NONLINEAR TIMESTEP_END TIMESTEP_BEGIN FINAL CUSTOM TRANSFER
Description:The list of flag(s) indicating when this object should be executed, the available options include NONE, INITIAL, LINEAR, NONLINEAR, TIMESTEP_END, TIMESTEP_BEGIN, FINAL, CUSTOM.
mem_typephysical_memoryMemory metric to report.
Default:physical_memory
C++ Type:MooseEnum
Options:physical_memory virtual_memory page_faults
Description:Memory metric to report.
mem_unitsmebibytesThe unit prefix used to report memory usage, default: Mebibytes
Default:mebibytes
C++ Type:MooseEnum
Options:bytes kibibytes mebibytes gibibytes kilobytes megabytes gigabytes
Description:The unit prefix used to report memory usage, default: Mebibytes
report_peak_valueTrueIf the postprocessor is executed more than once during a time step, report the aggregated peak value.
Default:True
C++ Type:bool
Options:
Description:If the postprocessor is executed more than once during a time step, report the aggregated peak value.
value_typetotalAggregation method to apply to the requested memory metric.
Default:total
C++ Type:MooseEnum
Options:total average max_process min_process
Description:Aggregation method to apply to the requested memory metric.

Optional Parameters

allow_duplicate_execution_on_initialFalseIn the case where this UserObject is depended upon by an initial condition, allow it to be executed twice during the initial setup (once before the IC and again after mesh adaptivity (if applicable).
Default:False
C++ Type:bool
Options:
Description:In the case where this UserObject is depended upon by an initial condition, allow it to be executed twice during the initial setup (once before the IC and again after mesh adaptivity (if applicable).
control_tagsAdds user-defined labels for accessing object parameters via control logic.
C++ Type:std::vector
Options:
Description:Adds user-defined labels for accessing object parameters via control logic.
enableTrueSet the enabled status of the MooseObject.
Default:True
C++ Type:bool
Options:
Description:Set the enabled status of the MooseObject.
force_preauxFalseForces the GeneralUserObject to be executed in PREAUX
Default:False
C++ Type:bool
Options:
Description:Forces the GeneralUserObject to be executed in PREAUX
outputsVector of output names were you would like to restrict the output of variables(s) associated with this object
C++ Type:std::vector
Options:
Description:Vector of output names were you would like to restrict the output of variables(s) associated with this object
use_displaced_meshFalseWhether or not this object should use the displaced mesh for computation. Note that in the case this is true but no displacements are provided in the Mesh block the undisplaced mesh will still be used.
Default:False
C++ Type:bool
Options:
Description:Whether or not this object should use the displaced mesh for computation. Note that in the case this is true but no displacements are provided in the Mesh block the undisplaced mesh will still be used.

Advanced Parameters

Input Files

modules/stochastic_tools/examples/batch/transient.i
test/tests/samplers/distribute/distribute.i
test/tests/postprocessors/memory_usage/print_memory_usage.i
modules/stochastic_tools/examples/batch/full_solve.i

modules/stochastic_tools/examples/batch/transient.i

[Mesh]
  type = GeneratedMesh
  dim = 1
[]

[Problem]
  kernel_coverage_check = false
  solve = false
[]

[Distributions]
  [uniform]
    type = UniformDistribution
    lower_bound = 1
    upper_bound = 9
  []
[]

[Samplers]
  [mc]
    type = MonteCarloSampler
    n_samples = 10
    distributions = 'uniform uniform'
  []
[]

[Executioner]
  type = Transient
  num_steps = 10
[]

[MultiApps]
  [runner]
    type = SamplerFullSolveMultiApp
    sampler = mc
    input_files = 'sub.i'
    execute_on = 'INITIAL TIMESTEP_END'
    mode = batch-restore
  []
[]

[Transfers]
  [runner]
    type = SamplerParameterTransfer
    multi_app = runner
    parameters = 'BCs/left/value BCs/right/value'
    to_control = receiver
  []
  [data]
    type = SamplerPostprocessorTransfer
    multi_app = runner
    vector_postprocessor = storage
    postprocessor = average
  []
[]

[VectorPostprocessors]
  [storage]
    type = StochasticResults
  []
[]

[Postprocessors]
  [total]
    type = MemoryUsage
    execute_on = 'INITIAL TIMESTEP_END'
  []
  [per_proc]
    type = MemoryUsage
    value_type = "average"
    execute_on = 'INITIAL TIMESTEP_END'
  []
  [max_proc]
    type = MemoryUsage
    value_type = "max_process"
    execute_on = 'INITIAL TIMESTEP_END'
  []
[]

[Outputs]
  csv = true
  perf_graph = true
[]

test/tests/samplers/distribute/distribute.i

[Mesh]
  type = GeneratedMesh
  dim = 1
[]

[Problem]
  kernel_coverage_check = false
  solve = false
[]

[Samplers]
  [sampler]
    type = TestSampler
    num_rows = 10000000
    num_cols = 1
  []
[]

[Executioner]
  type = Steady
[]

[Postprocessors]
  [total]
    type = MemoryUsage
    mem_units = 'bytes'
    execute_on = 'INITIAL TIMESTEP_END'
  []
  [per_proc]
    type = MemoryUsage
    value_type = "average"
    mem_units = 'bytes'
    execute_on = 'INITIAL TIMESTEP_END'
  []
  [max_proc]
    type = MemoryUsage
    value_type = "max_process"
    mem_units = 'bytes'
    execute_on = 'INITIAL TIMESTEP_END'
  []
  [test]
    type = SamplerTester
    sampler = sampler
    test_type = 'getLocalSamples'
  []

[]

[Outputs]
  csv = true
  perf_graph = true
[]

test/tests/postprocessors/memory_usage/print_memory_usage.i

[Mesh]
  type = GeneratedMesh
  dim = 2
  nx = 2
  ny = 2
[]

[Variables]
  [./u]
  [../]
[]

[Kernels]
  [./diff]
    type = Diffusion
    variable = u
  [../]
  [./dt]
    type = TimeDerivative
    variable = u
  [../]
[]

[BCs]
  [./left]
    type = DirichletBC
    variable = u
    boundary = left
    value = 0
  [../]
  [./right]
    type = DirichletBC
    variable = u
    boundary = right
    value = 1
  [../]
[]

[Adaptivity]
  [./Markers]
    [./uni]
      type = UniformMarker
      mark = REFINE
    [../]
  [../]

  # this marker will tag every element for refinement, growing the problem
  # exponentially with each timestep
  marker = uni

  # avoid a refine after the final step
  stop_time = 4.5
[]

[Postprocessors]
  [./physical]
    type = MemoryUsage
    mem_type = physical_memory
    value_type = total
    # by default MemoryUsage reports the peak value for the current timestep
    # out of all samples that have been taken (at linear and non-linear iterations)
    execute_on = 'INITIAL TIMESTEP_END NONLINEAR LINEAR'
  [../]
  [./virtual]
    type = MemoryUsage
    mem_type = virtual_memory
    value_type = total
    execute_on = 'INITIAL TIMESTEP_END'
  [../]
  [./page_faults]
    type = MemoryUsage
    mem_type = page_faults
    value_type = total
    execute_on = 'INITIAL TIMESTEP_END'
  [../]
  [./DOFs]
    type = NumDOFs
    execute_on = 'INITIAL TIMESTEP_END'
  [../]
  [./walltime]
    type = PerfGraphData
    section_name = "Root"
    data_type = total
  [../]
[]

[Executioner]
  type = Transient
  solve_type = 'NEWTON'
  petsc_options_iname = '-pc_type -sub_pc_type'
  petsc_options_value = 'asm      lu'

  nl_abs_tol = 1e-10

  num_steps = 5
  dt = 1
[]

[Outputs]
  csv = true
  execute_on = 'INITIAL TIMESTEP_END FINAL'
  perf_graph = true
[]

modules/stochastic_tools/examples/batch/full_solve.i

[Mesh]
  type = GeneratedMesh
  dim = 1
[]

[Problem]
  kernel_coverage_check = false
  solve = false
[]

[Distributions]
  [uniform]
    type = UniformDistribution
    lower_bound = 1
    upper_bound = 9
  []
[]

[Samplers]
  [mc]
    type = MonteCarloSampler
    n_samples = 10
    distributions = 'uniform uniform'
  []
[]

[Executioner]
  type = Steady
[]

[MultiApps]
  [runner]
    type = SamplerFullSolveMultiApp
    sampler = mc
    input_files = 'sub.i'
    mode = batch-restore
  []
[]

[Transfers]
  [runner]
    type = SamplerParameterTransfer
    multi_app = runner
    parameters = 'BCs/left/value BCs/right/value'
    to_control = receiver
  []
  [data]
    type = SamplerPostprocessorTransfer
    multi_app = runner
    vector_postprocessor = storage
    postprocessor = average
  []
[]

[VectorPostprocessors]
  [storage]
    type = StochasticResults
  []
[]

[Postprocessors]
  [total]
    type = MemoryUsage
    execute_on = 'INITIAL TIMESTEP_END'
  []
  [per_proc]
    type = MemoryUsage
    value_type = "average"
    execute_on = 'INITIAL TIMESTEP_END'
  []
  [max_proc]
    type = MemoryUsage
    value_type = "max_process"
    execute_on = 'INITIAL TIMESTEP_END'
  []
[]

[Outputs]
  csv = true
  perf_graph = true
[]

Implementation
Input Parameters
Input Files

Application Usage

Physics and Syntax

Application Development

Framework Development

MOOSEDocs

Infrastructure

Questions

Information and Tools