Esa_snappy crash on Linux during Sentinel-1 GRD processing

Hello! I am trying to run a Python script on Linux (SNAP 14) that uses the esa_snappy plugin to process a time series of Sentinel-1 GRD images. First each image is individually preprocessed by applying the following operators:

  1. Slice Assembly (if needed)
  2. Thermal Noise Removal
  3. Apply Orbit File
  4. Calibration
  5. Subset

After that, the DEM-Assisted-Coregistration operator is used to generate a stack of co-registered data, which are then multi-looked and geocoded using the Range-Doppler Terrain Correction operator. Finally, the subset operator is again applied. Unfortunately, my Python script crashes completely and I obtain the following error:

*** bit out of range 0 - FD_SETSIZE on fd_set ***: terminated

I have already tried the standard workarounds, such as increasing the terminal open files limit and implementing memory management in the script (saving intermediate .dim files to break the graph, using dispose(), gc.collect(), and Java System.gc()), but without success. In fact, despite these optimizations, the script keeps crashing. I get this exact same error even if I try to process just one single image.

Is there a way to fix this on Linux? The script runs without any problem on Windows.

Thank you in advance for your help.

Hi everyone, just a quick follow-up on this. I am still completely stuck with this problem in my workflow and haven’t managed to solve it yet. Has anyone else encountered a similar error or knows how to fix it? Thanks in advance!

Hi,

Could you identify at which processing step the crash occurs? Have you tried running the operators individually, using a single Sentinel-1 product, to determine which one triggers the error?

Also, could you try running the failing processing step directly with SNAP GPT instead of esa_snappy? This might help determine whether the issue is specific to esa_snappy or also occurs when using SNAP GPT.

Thanks!