FAQ

1. Why aren’t live logs showing up in the Web Console?

FDS behaves differently from most simulators: it writes all logs to stderr instead of stdout. Since the Web Console only streams stdout, no logs will appear there during the simulation.

To view the logs while the simulation is running, you can:

• Open the stderr file from the live storage.
• Use the CLI command inductiva logs <task_id> to stream logs in your terminal.

Once the simulation finishes, both stdout and stderr files will be available for viewing in the web console.

2. Why can’t I use more n_vcpus for my simulation?

FDS parallelizes simulations by assigning each mesh to a separate vCPU. If your simulation defines only a single mesh, it can only use one vCPU. Attempting to use more will result in an error.

For example, the following setup allows only 1 vCPU:

&MESH IJK=10,10,10 XB=-0.3,0.7,-0.4,0.6,0.0,1.0, MULT_ID='mesh' /

To utilize multiple vCPUs, define multiple meshes, like so:

&MESH IJK= 32,32,144, XB=14.0,18.0, 5.25, 9.25, 0.0,19.0 /
&MESH IJK= 72,21, 72, XB= 0.0,18.0, 0.00, 5.25, 0.0,19.0 /
&MESH IJK= 52,18, 72, XB=14.0,27.0, 9.25,13.75, 0.0,19.0 /
&MESH IJK= 56,34, 72, XB= 0.0,14.0, 5.25,13.75, 0.0,19.0 /
&MESH IJK= 36,37, 72, XB=18.0,27.0, 0.00, 9.25, 0.0,19.0 /

This setup allows the simulation to run using 5 vCPUs.

3. How does Fire Dynamics Simulator (FDS) use parallelization?

FDS benefits significantly from parallelization and supports two main methods:

• MPI (Message Passing Interface): This method divides the simulation domain into multiple meshes, with each mesh handled by a separate MPI process running in parallel.
• OpenMP: This enables shared-memory parallelism within each mesh by using multiple threads, improving performance on multi-core systems.

Still can't find what you're looking for? Contact Us