Processing time increasing after each run

Microwave Toolbox
#1

I’m processing some S1 GRD products using several GPT operators in sequence (the Apply Orbit step has already been run):

  1. Calibration
  2. Speckle Filter
  3. Terrain Correction

These are run using three gpt calls. Initially the combined processing time was around 15 minutes. However, when I decide to repeat the processing this time increases; if I keep repeating the processing it just increases further. I’ve encountered run times of 45 mins, 80 mins, up to nearly 3 hours. Each time I run the processing I am saving the DIM files in a new directory .

Setup:
SNAP 6.0
Debian Stretch
6 cores
16GB RAM
-Xmx10G
Tile cache size 1024
Tile size 512
Parallelism 6

I haven’t yet been able to test this on another machine but was wondering if anyone else has come across the same behavior. Perhaps there’s something I’m missing?

#2

Some times I faced up same problem, while using the same .xml, I think you could try to save the xml of new data in different name and place, might be help.

#3

Thanks for your reponse. Each time I run the processing I am saving the all data in a new directory. I’ll add this detail to my post.

#4

I meant the .xml as well, but one more thing to your first post, apply orbit! ,

Did you check up the update of your SNAP?

Also, try to exit the console and start a new one in your new process,

#5

Well spotted - I forgot to mention that the Apply Orbit process has already been run.

I’ll check for any updates that are available and see if exiting the console helps.

#6

I created a graph with some custom parameters to fulfill my requirements. Running these processes as a graph (as opposed to separate operators) drastically changed the run time. In addition, repeated processing seems to take the same amount of time; I am now getting a run time of 6 minutes.

I hadn’t been aware of the extent to which you can use custom parameters in graphs. My main reason for going with three separate calls to the operators was to provide additional operator-specific arguments. I found the Rice detection with Sentinel-1 using SNAP GPT RUS Webinar and step forum post (credit to @marpet ) very helpful.

I’m still not sure why using the operators was resulting in the increasing processing times but I have enough of a resolution to progress so I will flag as solved.

1 Like