Fallout Definitions
A fallout definition is the rule that turns a raw failure into a handled fallout. It names the error to match, the retry policy to apply, and whether an operator may edit the payload before retrying. Without a matching definition a failure is still recorded, but nothing retries it automatically.
Before you begin
- The error you want to match exists as an error definition, or you are ready to create one inline.
- A retry configuration exists and is active. Only active configurations are offered in the form at application scope.
Choose the scope
The form differs by where you open it, and the two are not interchangeable.
| Scope | Open from | Matches on |
|---|---|---|
| Application | The application's Configuration tab > Fallouts > Fallout Definition | A workflow (Service Name), a Step Name, and an Error Definition |
| Organisation | Settings > Fallout Definition | An Error Code only, in any application |
Use application scope when the handling depends on where the failure happened. Use organisation scope for an error that should be handled the same way everywhere. The application-level list also shows organisation-level definitions while the Organization Scope switch beside the search box is on.
Create a definition
- Select Add Fallout Definition. The Configure Fallout drawer opens.
- At application scope, identify the failing workflow in Service Name, then select
the failing step in Step Name and, optionally, the step to resume from in
Retry From Step. Both step lists come from the workflow you named, and the
workflow's
Endstep is excluded. - Select the error. At application scope the field is Error Definition; at organisation scope it is Error Code. The list shows each error's code with its description on hover, a globe icon on organisation-scoped entries, and an eye icon that opens the definition read-only. The last entry, Create Error Definition, opens the error drawer without leaving this form.
- Select the Retry Policy. The same list shape applies, and the last entry, Create Retry Policy, opens the retry drawer inline.
- Leave Retry Channel as it is. It is fixed to
HTTPand cannot be changed. - Enter a Name and a Description. Both are required.
- Switch on Allow data modify if an operator should be able to edit the record's payload before a manual retry.
- Switch on Active. A definition created from this page starts inactive; one created with Define Fallout from a live fallout starts active.
- Select Create.
The form has no alert field. Alert definitions cannot be attached to a fallout definition in this release; see Alert definitions for how alerting is configured instead.
Define a fallout from a live failure
The quicker route, and the one that fills the form for you. Open a fallout under Global Fallout or the application's Fallouts tab, and select Define Fallout in its details. The drawer opens with the service, step, and previous step pre-filled, and with the failure's error code selected. If no error definition with that code exists yet, Engage creates one from the failure's own name, description, category, and severity before the form appears.
From the application's Fallouts tab this switches you to the Configuration tab, where the drawer lives. From Global Fallout it opens in place, at application scope when the fallout belongs to a provisioned application and at organisation scope when it does not.
Manage definitions
The table shows Name, Error Code, Status, Description, and Actions, with a Service Name column at application scope. The Status switch activates or deactivates a definition in place; Engage confirms with Fallout Definition activated successfully or the deactivated equivalent. Select Edit to reopen the drawer as Edit Fallout Definition and finish with Update; select Delete and confirm to remove it.
How you know it worked
Engage confirms with Fallout Definition created successfully and the definition appears in the table. Failures that match it from now on show that retry policy's attempts on their timeline, and an exhausted one waits in the fallout list with its status set for manual handling. See Retry rules.
When it does not work
| Symptom | Cause | What to do |
|---|---|---|
| Error code is required or Retry policy is required | A required select is empty | Choose a value, or create one from the last entry in the list |
| Name is required or Description is required | One of the two text fields is empty | Fill both |
| The retry policy you expect is missing from the list | It is inactive; application scope lists active configurations only | Switch it on under Retry Config |
| Step Name offers nothing | Engage has no workflow profile for this application, which is the case for fallouts from applications outside any manifest | Define the fallout at organisation scope instead |
| Failed to auto-create the detected error definition | Define Fallout could not save an error definition for the failure's code | Create the error under Error Definitions, then define the fallout again |
| Failed to create fallout definition | The save was rejected; the message names the reason when the service gives one | Correct the named field and select Create again |
Next steps
Retry rules explains what the definition does once a matching failure arrives.