A workflow can start another workflow synchronously. The called workflow runs as an integral part of the calling workflow's run, and runs in the same memory space as the calling workflow. The calling workflow starts another workflow, then waits until the end of the called workflow's run before it starts running the next element in its schema. Usually, you call a workflow synchronously because the calling workflow requires the output of the called workflow as an input parameter for a subsequent schema element. For example, a workflow can call the Start virtual machine and wait workflow to start a virtual machine, and then obtain the IP address of this virtual machine to pass to another element or to a user by email.

A workflow can start a workflow asynchronously. The calling workflow starts another workflow, but the calling workflow immediately continues running the next element in its schema, without waiting for the result of the called workflow. The called workflows run with input parameters that the calling workflow defines, but the lifecycle of the called workflow is independent from the lifecycle of the calling workflow. Asynchronous workflows allow you to create chains of workflows that pass input parameters from one workflow to the next. For example, a workflow can create various objects during its run. The workflow can then start asynchronous workflows that use these objects as input parameters in their own runs. When the original workflow has started all the required workflows and run its remaining elements, it ends. However, the asynchronous workflows it started continue their runs independently of the workflow that started them.

To make the calling workflow wait for the result of the called workflow, either use a nested workflow or create a scriptable task that retrieves the state of the workflow token of the called workflow and then retrieves the result of the workflow when it completes.

A workflow can call a workflow but defer starting that workflow until a later time and date. The calling workflow then continues its run until it ends. Calling a scheduled workflow creates a task to start that workflow at the given time and date. When the calling workflow has run, you can view the scheduled workflow in the Scheduler and My Orchestrator views in the Orchestrator client.

Scheduled workflows only run once. You can schedule a workflow to run recurrently by calling the Workflow.scheduleRecurrently method in a scriptable task element in a synchronous workflow.

A workflow can start several workflows simultaneously by nesting several workflows in a single schema element. All the workflows listed in the nested workflow element start simultaneously when the calling workflow arrives at the nested workflows element in its schema. Significantly, each nested workflow starts in a different memory space from the memory space of the calling workflow. The calling workflow waits until all the nested workflows have completed their runs before it starts running the next element in its schema. The calling workflow can thus use the results of the nested workflows as input parameters when it runs its remaining elements.