Job Life Cycle During a Run
The following diagram shows the normal life cycle for an INVALID job.
| Status | Event | What keeps job in this status? |
|---|---|---|
| INVALID | If a job remains INVALID even after a run request, it is being blocked by a missing input (run vovforget -blocking). | |
| ↓ | A request is issued to perform a runtime tracing analysis for the job. | |
| SCHEDULED | If a job remains SCHEDULED forever, its inputs are not becoming VALID (investigate what happens to the inputs) or the FlowTracer server does not have access to sufficient resources (check the resource maps and the taskers). | |
| ↓ | All inputs for the job are VALID and all resources for the job are available and the job is dispatched to a vovtasker. | |
| RETRACING | ||
| ↓ | The job successfully fires on the vovtasker (this normally takes a few milliseconds). | |
| RUNNING | If a job remains RUNNING forever, a problem has occurred during execution. Either the vovtasker died, or the machine died, or more simply, the job is waiting for some user input. | |
| ↓ | The job completes. | |
| VALID or FAILED | If the job becomes VALID, all is Ok. If the job fails, you may want to investigate the cause of the failure. |