GRASS GIS in the Cloud

g.cloud is a tool for GRASS GIS to connect a personal GRASS session to a GRASS installation in a much faster cluster (high performance computing system). The general aim is to give the possibility to every user to perform high speed calculations, which are hard to compute on a personal computer by connecting to an external server. Today, in the new era of cloud computing, it is crucial to have cloud support in GRASS for a better and faster software performance and services.

The new extension is composed by two modules written in Python. The first one, g.cloud.prep checks if all program requirements are fullfilled on both the client and the server side. It also defines the needed variables and directory structures. The second module, g.cloud.launch, connects the current session to the cloud server, transfers scripts and data to the server and executes the remote job. Subsequently, it transfers back the resulting maps and data.

For the job execution, the script requirements are two, the user prepared GRASS GIS script which performs the calculations and server-side the qsub script from Grid Engine to launch the GRASS GIS script job.

The software requirements are GRASS GIS 7 and ssh/scp on the client side; while on the server side the Grid Engine software or compatible needs to be present to launch the GRASS jobs. To facilitate the usage, the g.cloud.prep module, if it does not find a GRASS GIS 7 installation on the server, it will download and install it automatically.

The new tool provides GRASS GIS users which potentially huge resources for massive data analysis and geospatial computing.