1
votes

I am running a simulation on a Linux server:

  1. Is it possible to run a simulation (experiment) on batch mode?
  2. The only document I found was this, but it says nothing about running experiments on batch mode.

Here an example of what I am trying to do:

  • I set a parameter variation experiment using a Latin hypercube sampling (240 parameter combinations):

  • I am saving text files of each run; I am running 10 replications per iteration.

I can run this using a Linux server, after running the application I get a window (XQuartz, I am using a Mac) from which I can manipulate and run the experiment:

./Anylogic/anylogic/anylogic
Edition: Personal
Starting Anylogic application
...
OS: linux64
Generate "MyModel"
Compile "MyModel"
Start HyperCube experiment
  • As far as I keep the window of the experiment open, the simulation will run. If I lose the connection with the server, the simulation will stop, unless I run this in a background mode (i.e., batch mode).

Any ideas? Examples?

1

1 Answers

0
votes

To run AnyLogic (multi- or single-run) experiments in 'batch', the normal method is to export the experiment to a standalone Java application, which can then be run as needed from the command line. This is available only in the Professional edition; looks like you are using the Personal Learning Edition (PLE).

Otherwise, AnyLogic is fundamentally a 'client-based' application. You might find some clever tricks to run it remotely as you suggest, but there is no 'intended' method, and you may be skirting along the edges of the license conditions.