Skip to content

feelpp/docker

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Docker 4 Feel++

Feel++ Env

What can you generate with this repository ?

Using docker and docker-compose, you will currently generate the following images:

  • feelpp/feelpp-env:latest

In-situ visulization with ParaView

In a nutshell

# This setup is valid if you are launching ParaView 
# and docker on the same computer.
# You must ensure that your ParaView version matches the one used in docker
# (It should be the latest available)

# Download Feel++ docker image
shell> docker pull feelpp/develop

From here, you have 2 ways to proceed:

  • Directly use the network of the host:
# Launch a pvserver
shell> pvserver
# Launch ParaView, connect to the pvserver and enable Catalyst in the interface 
shell> paraview

# Run the feelpp image
shell> docker run -ti --net=host feelpp/develop

# Run the application
dockershell> mpirun -np 2 feelpp_qs_laplacian_2d --config-file src/feelpp/quickstart/laplacian/qs_laplacian_2d.cfg --exporter.format vtk --exporter.vtk.insitu.enable 1
  • Use the port export feature from docker (safer regarding to security):
# Run feelpp image
shell> docker run -ti -P 11111:11111 feelpp/develop

# Launch tmux to multiplex terminals
dockershell> tmux
# Launch a pvserver in the first terminal
dockershell1> pvserver

# Launch ParaView and connect to the pvserver
shell> paraview

dockershell2> mpirun -np 2 feelpp_qs_laplacian_2d --config-file src/feelpp/quickstart/laplacian/qs_laplacian_2d.cfg --exporter.format vtk --exporter.vtk.insitu.enable 1

Full description

The dev-env image builds the latest stable version of ParaView with Catalyst enabled for In-Situ visualization. The develop image uses the dev-env base image and enables In-Situ visualization.

To use In-Situ visulization with a Feel++ application that you build in a container based on the develop image, you first have to configure the container to export ports.
The first time you launch the container with docker run, you must use the -p option to redirect a container port to the host machine. The option syntax is as follows in the simple case: -p hostPost:containerPort. In this case, you will redirect the containerPort port to the hostPort port.
We advise you to use a pvserver to use In-Situ. The default port pvserver uses is 11111, thus to export this port when running the container, you can do:

docker run -ti -p 11111:11111  feelpp/develop

Useful tip: the feelpp/dev-env image packages the tmuxand screen terminal multiplexers, so you can launch several virtual terminals inside the container.

Once you have launcher the container with the exported port, you can launch a pvserver inside the connect and setup your ParaView to use Catalyst and connect to the pvserver, by specifying the host and hostPort. You can then launch sample Feel++ apps and use In-Situ visualization. For example:

mpirun -np 2 feelpp_qs_laplacian_2d --config-file src/feelpp/quickstart/laplacian/qs_laplacian_2d.cfg --exporter.format vtk --exporter.vtk.insitu.enable 1

If you encounter problems regarding infiniband on supercomputers, you can add the following option to mpirun to use tcp: -mca btl tcp,self

For more information about In-Situ visualization in Feel++, please refer to https://github.com/feelpp/feelpp/wiki/In-Situ-Visualization