Introduction to the hybrid runner

Deploy your logic on-premises, private clouds, or personal devices. Spark leverages WebAssembly (Wasm) technology as a secure sandbox to execute near-native code, converting Excel logic into efficient C++ code, packaged as WebAssembly modules.

Key features

1. Privacy and security: Spark ensures the protection of sensitive data, making it an ideal choice for applications handling personally identifiable information (PII). The WebAssembly sandbox provides a secure environment for executing your logic, safeguarding your data at every step.

2. On-premises scalability: Take control of your computing resources with on-premises deployment. This option allows you to scale and optimize your infrastructure according to your specific needs.

3. Low-latency performance: Get fast, low-latency performance with the Neuron runner.

4. Telemetry tools: Support for OpenTelemetry to enable logging of service status and API calls.

How it works

nodegen-server is a Docker image that runs containerized applications for executing Wasm modules generated by Spark.

nodegen-server acts as a worker and does not handle user authentication or authorization. It does not store any sensitive data during its runtime, including when executing models or retrieving WebAssembly modules. You're responsible for securing the server and implementing the necessary authentication and authorization mechanisms, such as a proxy or API gateway.

Use the ImpEx APIs to export Wasm modules from Spark. These modules, along with the nodegen-server , can be neatly packaged into Docker containers. This process is designed with continuous integration/continuous delivery (CI/CD) in mind, ensuring a streamlined and automated deployment process.

Always use the latest version of the runner available. If you are not using the latest version of the runner, you may at risk of vulnerabilities or encounter issues when used with newer versions of Neuron.