Run Job (AI Hub) (Admin Tools)
SynopsisThis operator runs a given job in a given queue on an AI Hub.
This operator takes a process stored on an AI Hub and inserts it as a new job in the selected execution queue. This is equivalent to selecting the "Run Process on AI Hub" option in the Process menu in Studio, but uses a different API.
This operator is similar to the Schedule Process operator, which executes one job at a time. However, with the Schedule Process, the process containing the operator must be on the same AI Hub as the location of the process it is instructed to run remotely. In contrast, the Run Job (AI Hub) operator can run jobs on remote AI Hubs as well as on the AI Hub where Run Job (AI Hub) is executed. Note that the process selected for execution needs to be available on all remote AI Hubs.
Note that this executed only when it arrives at the top of the queue and there is an available job container available for execution. If there are no available job containers when the job reaches the top of the queue, the AI Hub will wait until one becomes available.
The operator will not wait for the execution to finish, nor will it throw an error if the process fails. To retrieve more information about the execution status or log details of the job, use the Get Jobs (AI Hub) and/or the Read Log (AI Hub) operators.
- con (Connection)
An AI Hub connection object with sufficient information and credentials (server_url, client_secret, and refresh_token) to access the AI Hub API.
The AI Hub connection object from the input port, delivered as a passthru.
- exa (Data Table)
An ExampleSet containing information about the job.
- location The location of the process to be run, e.g. /home/username/myprocess. Range:
- queue name Name of the queue you want to submit the jobs to. Range:
- macros This is an expert parameter. It defines macros for this process. Range:
- fail_on_error This is an expert parameter. If selected, the operator fails on any error. If not selected and a process fails, the operator returns an ExampleSet with the error message from the log, if available. Range: