SNAP Installation/Configuration


Im new to SNAP and Linux as well. We’ve installed SNAP in our Linux machine using the installer at Since we have access to high performance computing services, we would like to optimize the use of SNAP for our image processing tasks. One colleague of mine noticed the snap.conf and files and thought that maybe we can edit the configuration for SNAP to improve its performance. However, since we are still exploring SNAP, there are many things which we don’t fully understand yet. For example, there were settings which seem applicable only in Engine mode. What is the difference between Desktop mode and Engine mode? I’ve tried searching the documentation and found out that those are two subsystems of SNAP. Are they both installed by using the installer I mentioned above? How do we know that we are in Engine mode? Also, if you can help us point out the exact settings that we need to change that would be helpful. Thanks.

P.S. Here are the specs of our high performance computing service:
Total number of nodes: 48
Superserver 2027PR (48 x Intel ® Xeon ® CPU E5-2697 v2 @ 2.70GHz)
256 Gigabytes of RAM
2 x 10 Gbps ethernet

Hi and welcome to the SNAP community.

You are right, SNAP is split into two subsystems. The engine and the desktop.
The engine is the basis it runs in desktop mode but also from the command line. If you use SNAP as API you should start the engine on your own (Java: Engine.start();). In Python with snappy it would be started for you.
You can have a look at our wiki. There is a listing of properties for configuration. Then there is also the gpt.vmoptions file (next to the file). Here you can change the memory settings for gpt. I guess gpt is what you want to execute on your machine.

Thanks for the response! I’ll check out that link and gpt.vmoptions as well. So far, we’ve been using gpt in our batch processing tasks. I’m having a hard time using snappy since we’re using CentOS 6 with glibc 2.12 but snappy requires glibc 2.14. Similar cases with:


Forums say that glibc 2.12 is core part of the operating system ( and glibc cannot be updated safely ( Upgrading the operating system isn’t an option for us, so that’s why we’re stuck with gpt for now.