-
Notifications
You must be signed in to change notification settings - Fork 5
Using UCVM Docker Image
The SCEC UCVM developers are using container technology called Docker to distributed UCVM software. When distributed as a Docker image, UCVM users do not need to compile the UCVM software on their computer before using it.
One advantage of UCVM Docker images is they can be run on Mac, and Windows computers. This makes it possible for researcher with Mac or Window laptops (or servers) to run UCVM on their computer.
Typically, Docker images are useful for small to moderate scale UCVM queries up to 1M or so points. Larger scale queries, such as large simulation meshes, will probably still require UCVM source code installation on a Linux system.
Most registered CVM's are available as standalone UCVM Docker Images. The current list of models, using their UCVM abbreviations:
- cvms : CVM-S4 Southern California 3d Model - Magistrale/Kohler/Clayton/Graves
- cvmh : CVM-H v15.1 Southern California 3D Model - Harvard
- cvms5 : CVM-S4.26 - Chen/Jordan 3D tomography-based 3D model
- cca : CCA06 - Lee/Chen/Jordan Central Calfornia tomography-based 3D model
- cvmsi : CVM-S4.26.M01 - CVM-S4.26 with GTL properties from CVM-S4 restored to the top 350m
- cencal : USGS Central and Northern California 3D models
- albacore : ALBACORE - Offshore southern California 3D model derived from ocean-bottom seismometers
- cvlsu : Coachella Valley - 3D model developed using Salton Seismic Imaging Project (SSIP) dense seismic array data
- ivlsu : Imperial Valley - 3D model developed using Salton Seismic Imaging Project (SSIP) dense seismic array data
- wfcvm : Wasatch Front Utah Model - 3D model for Wasatch Front and Salt Lake City region.
To run Docker images on your laptop, users need to download the Docker client software from Docker and install it on their computer.
They will also need a Dockerhub account. Dockerhub
Users must start the Docker client on their laptop and wait until it has finished initializing. They then can open a terminal window on their computers to continue the installation.
Containers provide only termporary file storage. Files created in the container will be lost once the container is deleted.
So, users will want to transfer selected files between the Docker container and the laptops file system.
In the UCVM docker, files are exchanged by putting them in a "shared" directory that is visible and accessible to the container and the laptop user.
In UCVM, the shared directory is called target.
The target directory is where users put large input files (e.g. mesh.txt) or where ucvm writes large output files.
/Users/maechlin/ucvm_docker
cd /User/maechlin/ucvm_docker mkdir target
$git clone -g target --single-branch sceccode/ucvm.git
cd ucvm/docker
$ run_ucvm_docker.sh
- Overview and Introduction
- Installation
- Registered CVMs
- UCVM Configuration File
- Setting the UCVM Environment
- Basic UCVM Commands
- List of Commands
- Tutorials
- UCVM-Docker
- References
- UCVM Query Parameters
- Model Specific Query Behavior
- Model Specific GTLs
- System-specific Builds
- Model References
- Concepts
- Test Cases
- Earlier Documentation
