GPT stop without crash or error message in Ubuntu docker environment

Dear all,

we built up a SNAP model graph with these nodes:

  • Read (original S1SLC)
  • Calibration
  • ThermalNoiseRemoval
  • TOPSAR-Deburst
  • Multilook
  • Speckle-Filter
  • Terrain-Correction (using external DEM)
  • SRGR
  • GRD-Post
  • Write (to BigTIFF)

It runs well in SNAP GUI under Windows. We are struggle with this model in GPT under Ubuntu docker environment, because on some SLC images it runs without any problem, but sometimes it’s stopped around 30% (for example using this image: S1A_IW_SLC__1SDV_20191107T044509_20191107T044536_029800_0365C6_C397.SAFE). We don’t get any error message on standard output, and the GPT don’t crash. We waited for hours, but it’s still on 30%. It seems, the GPT has enough RAM/CPU/Storage. We tried several times using higher Xmx and -c, like “40G Xmx –c 30G”, but the problem is still occur. Any suggestion in order to solve this problem?

Many thanks in advance!
András