When defining packages rapyuta.io allows the developer to directly reference source code from a VCS like GitHub or Bitbucket. The platform then takes the responsibility of cloning, building and delivering the software across the cloud and device.
Under the hood, rapyuta.io solves the hard problems of native arm compiles, software versioning, artifact delivery, transactional upgrades and provisioning.
With all these features backed by a complete API, it makes it possible for the developer to automate the flow from git to operations and integrate it with existing CI/CD systems and QA processes.
rapyuta.io builds ROS and non-ROS packages using various build strategies. The Build Engine selects an appropriate build strategy based on whether the package contains a git repository with Dockerfile or without it or a docker image.
The goal of each build strategy is to generate a running docker container at the end of the package creation process.
This strategy builds source code into a docker image. The source code is usually stored in a git repository. If it is a private git repository, you need to add a source secret to access the repository contents. rapyuta.io uses ROS Builder, a subset of catkin build, to build source code into a docker image. If a package contains a ROS component, the component’s ROS version can be Kinetic or Melodic. When building the source code, the ROS Builder takes into account its corresponding component’s ROS version.
Set the Executable Type as Git and provide the URL address of a git repository. Suppose you want to add the address of a git repository say https://github.com/rapyuta-robotics/io_tutorials, where io_tutorials is the project folder that contains the source code on the master branch and is hosted on GitHub.
If you want to add source code located on a different branch, say io_turtlesim_qos of the same project, your git repository URL will look like: https://github.com/rapyuta-robotics/io_tutorials#io_turtlesim_qos
The ROS Publisher Subscriber walkthrough is an example of the source code build strategy.
rapyuta.io supports cross-compilation of source code to be able to run on devices with arm64, arm32 CPU architectures.
rapyuta.io supports cloning and fetching from repositories that deal with large files (as large as a couple of GB in size) with Git Large File System (LFS) extension. If you want to use Git LFS with a private git repository, you must select SSH authentication while adding a source secretbecause the LFS extension will fail with the Basic authentication, and the corresponding source secret will not be applied to the private repository.
This strategy builds a Dockerfile into a docker image. The Dockerfile is usually saved in a git repository. If it is a private git repository, you need to add a source secret to access the repository contents.
You may explicitly specify the absolute path of the Dockerfile, or the root of the git repository is set as the location of the Dockerfile.
The Basic Web Application walkthrough is an example of the dockerfile build strategy.
This strategy uses a pre-built docker image. The docker image is usually stored in either a public docker registry (i.e., Dockerhub) or a private docker registry. You need to add a docker pull secret for rapyuta.io to access a private docker image.
The maximum size of a docker image for cloud deployment is 10GB.
If you are going to deploy a docker image onto a device, ensure that the CPU architecture of the device is compatible with that of the image being deployed. You may use the Build Engine for compiling for a different target architecture.
The Hardware Interfacing walkthrough illustrates the docker image strategy.