You may set up multiple tasks or services, however, if Allow Multiple Services to Execute Concurrently is not enabled, only one task will execute at a time. Essentially, additional tasks detect that another task is active, and immediately completes the execution. Tasks will continue to repeat this check based on their execution window, and execution interval configuration. Configuring multiple services is a way to achieve high availability. If one server or service fail, another takes over after 60 seconds.
You can also set up tasks for horizontal scaling, which allows you to configure the HL7 Back-end Processor in both Server and the Client modes. This is accomplished by enabling Allow Multiple Services to Execute Concurrently on all tasks. The first service that is run sets itself as the Server. Any subsequent services starting afterwards label themselves as Clients only. A Server manages jobs which can safely be processed concurrently by Clients. A task that sets itself up as a Server will also process jobs (essentially, it can perform both the function of a Server and a Client). If the Service for the Server goes down for some reason, the Clients try to detect this, and then try to upgrade themselves to be the Server (in the case of multiple clients, this would be the first client to make that switch). For best practices, the separate task groups and separate services are mandatory. The timing of each task must use the same configuration for all tasks that are set up. When setting up a task for horizontal scaling, you must:
- Enable the Allow Multiple Services to Execute Concurrently option.
- Set up two separate Back-end processor tasks, each with their own unique service and in their own respective task groups with the same Port and Certificate Thumbprint (if using secured connections).