The future of reproducible and replicable computing in the academy is directly tied to strong digital research infrastructure that can support the needs of these workflows. Often this means providing high performance computing power and or controlled access to large datasets through virtual machines or containers that are the reproducible objects for work undertaken. Whether we are working with provisioner files which can help define and run an installation process to “provision” or install everything necessary for a process to run, to more comprehensive environment managers, containers, and virtual machines. Many tools exist to allow you to take a snapshot of the context in which you are working.
The use of container tools such as Docker or Singularity, environment managers such as Vagrant, and dedicated provisioners such as Ansible can go far in helping to make the environment in which work occurred reproducible. We will touch on this further in the Open Software section.
Dig Deeper
Learn more about using containers in research and digital research infrastructure:
- Read about using Docker in research
- Read more about Data Capsules
- Read about containers specialized to High Performance computing.
- Learn about digital research infrastructure in Canada by reading “What is Digital Research Infrastructure“
Explore virtual research environments, a tool for the replication of these environments, and discussion about how simple or minimal approaches can remove some of the barriers to preservation of outputs.
- Example of virtual research environment architecture: PARTHENOS
- Reproducible Zipping with ReproZip
Finally, read about a related approach to replicable computing: Minimal Computing.