Implementation details
The Helm chart handles the creation of resources automatically, resulting in a complete infrastructure for each dedicated runtime pod.
When the dedicated runtime pod feature is enabled, it creates the following Kubernetes resources
for each dedicated configuration:
- Deployment: A separate pod deployment for each configured dedicated section, with the specified resources and configuration
- Service: A dedicated Kubernetes service for each deployment to enable internal communication
- Ingress Rule: A specific ingress rule to route external traffic to the dedicated pods based on the defined paths
The regular Decision Server Runtime pods and services are still created alongside these dedicated resources. Any requests with paths that do not match the dedicated runtime paths are automatically routed to the default Decision Server runtime.
When deployed, the Helm chart does the following actions:
- Creates separate pod deployments for each configured dedicated section.
- Routes requests matching the specified paths to their respective dedicated pods.
- Configures the main Operational Decision Manager runtime to handle the routing logic for any unmatched paths.
- Maintains the ruleset isolation across different pods.
- Applies the specified resource constraints to each dedicated pod.
- Uses the specified webConfigRef and xuConfigRef to customize the runtime behavior and execution unit optimization.