View on GitHub

RayScale - Ray Tracer

Microservices based ray tracer designed to run in containers for scaled & batch operation

RayScale Controller

Configuration

All configuration is done via environmental variables. When running locally dotenv (.env) files are the simplest way to modify the configuration, a sample .env.sample file is provided which can be renamed. When running as a container you must provide the environmental variables using what ever interface the container runtime provides (command line, YAML, ARM template etc)

Name Default Notes
PORT 9000 The port the server listens on
HEALTH_CHECK_INTERVAL 90 How often in seconds, to ‘ping’ tracers to check they are healthy and online. Setting lower than 60 sec with remote tracers you might trip over TCP timeouts (often 60 secs). Setting lower than 5 seconds is a bad idea.
DIR_JOBOUT ./dist/../jobs Where to store jobs, folder will be created if it doesn’t exist. The default will be folder above the dist folder with compiled JS

Controller API

Route Method Calls Notes
/api/status GET getStatus() Provide current status of the controller, active job, tracers online etc
/api/jobs GET listJob() Lists completed and active jobs, by examining the jobs output folder
/api/jobs POST startJob() Start a new job, POST body must contain a YAML job definition, and have content-type of application/x-yaml
/api/tracers POST addTracer() When a tracer is started, it registers with the controller using this
/api/tracers GET listTracers() List which tracers are online
/api/tasks/{taskId} POST taskComplete() Called by tracer on completion of its task, normally binary image data

Data Structures

Job Input

name:   string;     // Job name, no spaces
width:  number;     // Output image width
height: number;     // Output image height
scene:  Scene;      // Scene to be rendered