Controlling a processing
This page describes the actions available on a single Processing. They are reached from the actions menu on the processing in the Active processings panel of the Monitor console, and require the Admin role.
These actions affect one processing only. To stop or resume everything at once, see Executor control.
Contents
Show processing details
Opens the processing info dialog, showing the label, scenario, date, iteration, status, start and finish time, and Invoked by — the user who started it, or System when it was planned automatically.
The dialog also lists the parameter values resolved for this processing. This is the reliable way to see what a parameter evaluated to at planning time, which matters because parameters defined by database queries are evaluated once, when the processing is planned, and not again for each task. See Planning.
Pause and resume
Pause suspends one processing. No new task from it will be started; tasks already running continue to completion. Other processings are unaffected.
Use it when:
- a problem needs investigating and you want to stop the run advancing while you look
- a target system needs to be taken out of service briefly
- a corrective load has to be applied before the remaining tasks run
Resume returns the processing to normal execution.
While a processing is paused, the action menu on its tasks reports Processing not active and the task actions stay disabled. This is expected — resume the processing first.
Add tasks
The set of tasks in a processing is fixed when the processing is planned. Tasks created or enabled afterwards are not part of it.
Add tasks resolves this without re-planning. It is available while the processing is Running or Paused, and opens a dialog listing the tasks that could be added but are not currently included. Search by name, tick the ones required and confirm with Add.
The typical case is a task added to the definition after the day's processing had already been planned — for example a new extract that has to run today rather than from tomorrow.
Added tasks are subject to the same dependencies and resource limits as the rest.
Reopen
Reopen applies to a processing that has already finished. It is used when work has to be done again within the original processing rather than as a new run.
The processing moves to Reschedule status and waits for the executor to re-plan it. After re-planning:
- all tasks are in finished status
- the processing itself is left in Paused status, so that it stays active
The processing is deliberately left paused. Nothing runs until you decide what should. The normal sequence is to reopen, then use Repeat or Force repeat on the specific tasks that need to run again, then Resume the processing.
Reopen keeps the work attached to its original processing date and iteration. Choose it over reprocessing when the correction belongs to a run that has already completed and you want the history to reflect that.
Choosing between the options
| Situation | Action |
|---|---|
| Stop a run advancing while you investigate | Pause the processing |
| A task was added to the definition after planning | Add tasks |
| A finished run has to be partly redone, keeping its date | Reopen, then repeat the specific tasks |
| Selected tasks must run again over a range of dates | Reprocess |
| Nothing at all should run for a while | Pause the executor — see Executor control |
