david4096/mapd-core repository overview

⁠Docker and nvidia-docker Installation

Install Docker and nvidia-docker:

• https://docs.docker.com/engine/installation/⁠
  • Be sure to use Docker's repositories for the latest version of Docker.
• https://github.com/NVIDIA/nvidia-docker/blob/master/README.md⁠

To use Docker as a normal user, add the user to both the docker and nvidia-docker groups:

sudo usermod -aG docker username
sudo usermod -aG nvidia-docker username

⁠Building MapD container

The Dockerfile assumes a copy of the MapD tarball is in the same directory as the Dockerfile and is named mapd-latest-Linux-x86_64.tar.gz.

To build the container, run:

mv ../../mapd-3.0.0-*-render.tar.gz mapd-latest-Linux-x86_64.tar.gz
nvidia-docker build .

where ../../mapd-3.0.0-*-render.tar.gz is the path to the MapD tarball.

The container image id will be output on the last line of the build step. To assign a custom name and tag:

nvidia-docker build -t mapd/mapd:v3.0.0 .

which will assign the name mapd/mapd and the tag v3.0.0 to the image.

⁠Image layout

The tarball is extracted to /installs. The extracted tarball is also symlinked to /mapd.

The data directory is at /mapd-storage/data.

The config file lives at /mapd-storage/mapd.conf.

⁠Running MapD inside a container

nvidia-docker run -d \
  -p 9092:9092 \
  --name mapd \
  -v /path/to/mapd-storage:/mapd-storage \
  -v /usr/share/glvnd/egl_vendor.d:/usr/share/glvnd/egl_vendor.d \
  mapd/mapd:v3.0.0

This starts the MapD Core Database inside a container named mapd, and exposes the Immerse visualization client on port 9092..

Data will be persisted to the host directory /path/to/mapd-storage.

The /usr/share/glvnd/egl_vendor.d directory is required for rendering support when using recent NVIDIA GPU drivers.