CEA Codes forum

Hi Didier,

When I try to run an error study on my local Linux cluster, even with only 10 particles, 1 run, 1 step, and a single core, I get a segmentation violation error when I run my error study. The error log shows a repeated error. I am able to connect to the cluster fine -- I tried using "Test computers", and I have separately verified that twserver is running on the cluster and is listening to port 9093. The simulation itself runs fine on my local computer though (i.e. not within the context of an error study).

I'm not sure what the source of the error is, as it is a segmentation error. Any ideas? I will send you my simulation files separately.

Thanks,

Josh

I'm not sure if this is relevant to this issue, but I should note that I use TraceWin on a Windows computer, but my local cluster is composed of Linux computers. When I tried to run a simple test simulation using the TraceWin linux executable, it says it has limited functionality (i.e. only up to 20 elements), despite copying my tracewin.key and toutatis.key files in the executable directory. Do you think this is causing the segmentation error?

Josh

I figured out what was causing the segmentation errors...it has nothing to do with Windows vs Linux.

The .sh batch files as written in the documentation contain two typos (see below). In both cases, the quote should be inside the closing parentheses, not outside. Can these be fixed?

Josh