The radanalyticsio/oshinko-webui repository provides a web-based console for managing Apache Spark clusters within an OpenShift environment. Its primary purpose is to simplify the deployment, monitoring, and lifecycle management of Spark clusters for users who operate within OpenShift projects. By offering a graphical user interface, oshinko-webui abstracts away much of the complexity involved in cluster management, making it accessible to both data scientists and developers who may not be deeply familiar with Kubernetes or OpenShift command-line tools.
The application is built as a Node.js service, packaged to run inside an OpenShift Pod. Once deployed, it allows users to create, update, and delete Spark clusters directly from their web browser. This is particularly useful in environments where rapid provisioning and teardown of Spark clusters are required for data processing tasks, experimentation, or development workflows. The web UI interacts with OpenShift resources using a service account with appropriate permissions, ensuring that users can manage Spark clusters securely within their own projects.
Installation of oshinko-webui is straightforward for both end-users and developers. For users, the process involves creating a service account and deploying the web UI template using the OpenShift command-line tool (`oc`). This setup ensures that the web UI has the necessary permissions to manage Spark resources. Developers interested in contributing to or customizing the web UI can set up a local development environment using Node.js (with recommended version 6.3.1), npm, and Bower for dependency management. The repository provides scripts and instructions for running the application locally, including setting up a proxy to communicate with the OpenShift API server.
Testing is a significant focus of the project, with support for both unit and end-to-end (E2E) tests. Unit tests can be run using Karma, while E2E tests are facilitated by scripts that automate the setup of test environments, including service accounts, templates, and necessary OpenShift resources. The E2E testing framework is flexible, allowing configuration through environment variables to accommodate different registry setups, secure web UI deployments, and custom resource files. This ensures that the application can be reliably tested in a variety of OpenShift configurations, including those with integrated or external Docker registries.
The oshinko-webui project is designed to be robust and maintainable, with automated builds and vulnerability checks integrated into its development workflow. It is distributed as a Docker image, making it easy to deploy and update within OpenShift clusters. The web UI is intended to work seamlessly with other tools in the radanalytics.io ecosystem, such as the oshinko-cli, providing a comprehensive solution for Spark cluster management on OpenShift.
Overall, oshinko-webui addresses the need for an intuitive, browser-based interface for Spark cluster management in OpenShift environments. It lowers the barrier to entry for users who need to run distributed data processing workloads, supports both production and development use cases, and provides a solid foundation for further extension and integration with other data analytics tools.
