Batch Simulator

Batch simulations can be setup by selecting New Batch menu after right-clicking on Simulations in the Project Panel.

The batch simulation enables the utilisation of multiple processors (where available), thus further speeding up simulation runtimes. The function automatically detects the number of processors available on the host system. Subsequent batched simulations will then be scheduled to run concurrently, using the next available processor. Multiple processors can also be utilised for the 2D Solver and the TUFLOW Link batch simulations.

When a New Batch menu is selected (from the right-click menu on Simulations in the Project Panel) a new window is displayed for the modeller to define the simulations to include in their batch run.

The approach is based largely on the one *.ief file per scenario principle, and these should be created before running the batch process.

The three basic steps to running batch simulations are:

  1. Create an ief file for each scenario (see Scenario data section)

  2. Enter the ief file names in the table

  3. Launch the batch simulation

To add a new Scenario file to the table, use the Add button, which opens a file dialog to select the *.ief file name and inserts it in a new row of the table. Alternatively, the button in the Filename column of the table may be used to access the file dialog.

Manipulation of the table can be performed by using the Remove button (removes current row) and the Up/Down Arrows (move currently highlighted row up/down). The context (secondary mouse button) menu also has options to insert rows above and below or to remove the current row, as well as to clear the entire grid and to select the *.ief file name.

The Scenario Title column (read-only) is automatically filled from the Event Title field in the corresponding *.ief file.

The Select column enables the user to select whether to run the simulation for the specified scenario on launching the batch simulation.

The status column (read-only) is only populated after a batch simulation has been performed during the current session and displays the run status of the corresponding scenario as follows:

Status

Meaning

Running...

Scenario currently running

No such file

Specified scenario file (*.ief) does not exist

Completed

Successful simulation; fully convergent

Completed (NC)

Successful simulation; periods of non-convergence

Failed (3)

Simulation failed mid-run; fully convergent until then

Failed (4)

Simulation failed mid-run; periods of non-convergence

Failed (5)

Simulation failed at start

Security error

Simulation failed mid-run due to dongle error (or similar)

Unselected

Scenario was not selected for simulation

Stopped

Status unknown1

Note:

  1. Status=Stopped does not necessarily mean an unsuccessful run; the status code is not received by the interface, resulting in this return value.

If the (default) Graphics-free simulation option is chosen, the alternative graphics-free computational engine is used; computationally, the two are identical, although a non-fatal interruption requiring user intervention occasionally occurs within the graphical version on certain machines. This would pose problems for unattended batch runs, hence the reason for the alternative version. Note that run-time graphics (see below) are not saved with the graphics-free version.

It is recommended that the Force Auto Shutdown box be checked when running a batch simulation, unless user intervention is desired. This option allows the majority of errors (if any occur) to be trapped before passing to the next scenario simulation. If the Graphics-free simulation option is not chosen, the run-time graphics are also automatically saved to a *.bmp file (with same root name as the results file). If Force Auto Shutdown is unchecked, the usual 1D Solver post-run "View File" dialog will appear at the end of each simulation, which requires user input before proceeding.

The Auto Postprocess check box allows simple Tabular CSV postprocessing to be performed after each simulation. If this box is checked, the Tabular CSV script file name must be entered in the adjacent box. The script file may be created within the Tabular CSV program by entering the required parameters, and choosing "Save Settings" from the context menu. The script file is generally only applicable to different simulations with the same node structure and run times as the original.

The multi-processor batching options provide the following controls to define how to utilise multiple processors:

The buttons towards the base of the window may be used as follows:

Log file

A historical log file is automatically written each time a batch simulation is undertaken, consisting of the simulation files and parameters (other than the defaults), time and date of the start and end of the run and the name of scenario files used for each scenario. The name of the log file is the same as that of the batch simulation file (*.ipf), with the extension .log appended (i.e. *.ipf.log). If the scenario file has never been saved, then the log file will be written to the Windows TEMP directory.

The log file may be viewed within the Project Data interface by selecting File | View Log File, or by using any text editor. Writing of the log file may be disabled by toggling from the Run | Write Log File menu item.