Deployment

A deployment with the OpenTalk Orchestrator requires configuration changes in the OpenTalk Controller and in the services that should be scaled.

Scaling a service with the orchestrator takes three steps:

1. Add the orchestrator section to the configuration of the proxied service.
2. In the controller.toml, replace the url and api_key of the proxied service with the orchestrator's proxy URL and API key.
3. In the orchestrator.toml, add the API keys of each proxied service to the services.keys configuration.

The following services can be scaled when configured with the orchestrator:

Service	Orchestrator endpoint
OpenTalk RoomServer	/roomserver/
OpenTalk Recorder	/recording/

Related component configuration

• Controller
• Recorder
• Roomserver

Running multiple Orchestrators

The orchestrator can run with multiple instances in parallel. For this to work, the orchestrator storage configuration needs to be set to redis and there has to be a load balancer that forwards requests to the orchestrator instances.