Error submitting FDTD simulations (mode source) on cluster




I have a simulations which is basically launching a mode source in a ridge SOI waveguide, to a W1 waveguide, and transmissions are collected. However, when I’m submitting this type of simulations to clusters, they cannot run and throw errors like:

[pod12b15:28529] Signal: Segmentation fault (11)

The same simulations can run properly on my local server via GUI. Also, if I change the mode source to a dipole source in the waveguide, they run properly on the cluster.

Can anyone let me know what I did wrong?

Thanks in advance.



To investigate further, can you provide more details or the following?

  1. submission script you are using for these simulations?
  2. the simulation file, that does not work in the cluster
  3. and the one that works?



Thank you so much, Lito!

Here are the files:

  1. (I changed the .sh into .txt for uploading)
    fdtd-run-8.17.txt (2.1 KB)

dy70.fsp (2.9 MB)

dy70_test.fsp (2.9 MB)

Thanks again!



From the files, you are using a PBS scheduler and using our example template.
How many nodes, cores/node, memory/node and time to run do you request when submitting the job?
When running the simulation how many # of processes and # of threads do you use?
Can you send a screenshot of the actual error message?



Thanks, Lito.

For number of process, I usually use 64 or 128. Memory and time requested is 500%, and there’s no problem submitting/starting the job.

The error message is in the file: (4.8 KB)

Let me know what other information you need.



I will forward this to our development team and check if this is a bug. We will get back to you as soon as possible.

Thank you for the updates.



Do you happen to know the hardware and operating system information / details / version for your cluster?



I’ve tried on both orcinus and cedar from westgrid. Their specs are as follows:



We will check if this is something with the MODE source simulations and get back to you. Thank you for your patience.



Can you provide the versions of FDTD running on the cluster and the versions to are using on your local workstations?

Thank you,



orcinus: 8.17.1157
cedar: 8.19.1466
local workstation: 8.18.1262



Please try to update to the latest version that is now available on our downloads page. There has been several bug fixes since the versions that you are using on your clusters.

Please let us know how this pans out.