Tapestry (Scientific Visualization as a Microservice)

Powered by Intel Rendering Framework

Tapestry is a platform for creating lightweight, web-based volume rendering applications at scale, for many users.

Requirements

• Docker (https://www.docker.com/get-docker)

Installation

Run ./tapestry.sh depend to fetch and install the Tapestry submodules.

Running ./tapestry.sh build will then build and install the Tapestry Docker image. You can use -j to specify the number of processes for building. Use -m to minify the Javascript internally.

Running the example

• To run the example, first download the data, the configurations, and the example app using ./tapestry.sh examples
• Second, run ./tapestry.sh run -c examples/configs/ -d examples/data -a examples/app
• Third, navigate to http://127.0.0.1:8080 in your browser
• tapestry.sh provides all of the management scripts needed for building and running. Run ./tapestry.sh -h for more options
• Since Tapestry uses Docker Swarm, to kill the running service, run docker service rm tapestry

Usage

To use Tapestry with your own page and datasets, you will need three things:

1. A directory with your datasets (currently, Tapestry supports raw single variable binary as well as NetCDF files)
2. A directory with one or more configuration files that point to the data. You can use the provided examples above as a starting point
3. An index.html with hyperimage and optionally, hyperaction tags

You can provide additional Tapestry options by editing tapestry/enchiladas/src/js/main.js after doing an initial build. You would also need to rebuild the image after any edits.

If you use Tapestry, please cite one or both of these two papers:

@article{raji2018scientific,
  title={Scientific Visualization as a Microservice},
  author={Raji, Mohammad and Hota, Alok and Hobson, Tanner and Huang, Jian},
  journal={IEEE Transactions on Visualization and Computer Graphics},
  year={2018},
  publisher={IEEE}
}

@INPROCEEDINGS {Tapestry2017,
    author    = "M. Raji and A. Hota and J. Huang",
    title     = "Scalable web-embedded volume rendering",
    booktitle = "2017 IEEE 7th Symposium on Large Data Analysis and Visualization (LDAV)",
    year      = "2017",
    pages     = "45-54",
    month     = "Oct",
    doi       = "10.1109/LDAV.2017.8231850"
}

More documentation can be found in the wiki.