Global Fallout
A fallout is an execution that failed and now needs an operator. Global Fallout lists every fallout in the organisation for the selected environment, across all applications, so an operator can work them without opening each application in turn. The same list, scoped to one application, is the Fallouts tab on Application details.
What Engage does with a fallout automatically, before it reaches this list, is decided by fallout definitions, retry configuration, and alert rules. See Managing fallouts.
Open it
Select Global Fallout in the sidebar. The sidebar lists Dashboard, Applications, Delivery Failures, and Global Fallout at the top, then the groups Fallouts, Organisation, and Account. The list is scoped by the environment in the header, Development or Production, and by the date range control in the toolbar.
Read the status cards
Status cards sit above the list and count the fallouts in the selected period.
| Card | Counts |
|---|---|
| Total Fallouts | Every fallout, whatever its status |
| New | Fallouts Engage has just received |
| Failed | Fallouts whose automatic or manual retry failed |
| Resolved | Fallouts that recovered after a retry, or that someone marked resolved |
Select a card to filter the list to that status; select Total Fallouts to clear the status filter. The cards keep showing the totals for every status while one is selected, and show a dash instead of a count while a search is active.
The list
| Column | Meaning |
|---|---|
| Timestamp | When the fallout was recorded. Sortable. |
| Status | The fallout's lifecycle status, in plain words such as New, Retrying, Auto Retry Failed, Recovered, Manual Retry Failed, Manually Recovered, or Resolved. Filterable. An Exhausted tag beside it means automatic retries have run out; the fallout can still be retried by hand. |
| Application | The application the execution belongs to. Filterable. |
| Workflow | The workflow that failed |
| Manifest | Configured or Not Configured: whether the application has a manifest. Filterable. A fallout without one can only be retried or restarted; see Choosing a retry method. |
| Service | The service the failing step called. Filterable. |
| Step | The workflow step that failed. Filterable. |
| Error Code | The error code, matched against error definitions. Filterable. |
| Error Message | The error text |
Application, Workflow, and Manifest appear only on this page; the application-scoped Fallouts tab omits them. Hover over a cell to see its full text in a tooltip, with a copy button that copies it. The toolbar offers a search box, Refresh, the date range, Use Query, and Expand to Fullscreen. An empty list reads No fallouts found for the selected period. Try adjusting the date or environment filters.
Find a fallout
- Search. Type in Search or paste a fallout ID to search the period's fallouts. Pasting a fallout ID and pressing Enter opens that fallout directly.
- Column filters. Use the filter control in a column header. Each active filter appears as a chip above the list; remove one with its chip, or select Clear all.
- Use Query. For several conditions at once, select Use Query. The Query Search dialog builds a filter from rules, each a Column (such as Fallout ID, Status, Application, Service, Step, Error Code, or Error Message), an Operator, and a Value. Select Rule to add another. In Match, choose All rules (AND) to require every rule or Any rule (OR) to accept any one. Select Search to apply, or Reset to start over. Queries run against every fallout in the period, not only the page on screen.
Starting a search clears the query and column filters.
Link to a fallout
A fallout's page has its own address. Opening a link that carries a fallout ID opens that fallout. If it belongs to the other environment, Engage switches environment and says so; if it belongs to another organisation, switch to that organisation from the organisation menu to act on it. A link that cannot be resolved reports Fallout not found, or you don't have access to it.
Retry several at once
Rows whose status allows a manual retry have a checkbox at the left.
- Select the rows to retry.
- Select Retry Selected (N). The count is the number of selected rows that can be retried.
- In the Retry N fallouts dialog, choose an Action: Retry re-runs each failed step from where it left off; Restart runs each workflow again from the start.
- Select Confirm Retry.
Engage reports Retry sent for N fallout(s) (or Restart sent for N fallout(s)) and refreshes the list. Bulk retry sends each fallout's data as recorded. To change the data, retry from a fallout's own page, or to start from a chosen step or a retry workflow, see Fixing a fallout.
To close the selected fallouts without re-running them, select Mark as Resolved (N), enter a Reason, and select Mark as Resolved. The reason is saved to each fallout's timeline.
Work one fallout
Select a row to open the fallout. Its page shows what happened, the data a retry sends, the steps, and a status timeline of every run, with the actions to retry it, mark it resolved, open it in Studio, or define a fallout definition for it. It is documented in Fixing a fallout; the retry options are compared in Choosing a retry method.
When it does not work
| Symptom | Cause | What to do |
|---|---|---|
| Retry is disabled | The fallout is not in a manual state; Engage is still retrying it automatically, or it is already resolved | Wait for the automatic retries to finish, shown by the Exhausted tag, then act |
| Retry Selected is disabled with rows selected | None of the selected rows is in a manual state | Select only rows whose checkbox is enabled |
| Retry failed | The retry request was rejected | Check the failed step in the fallout's Metrics view, then try again |
| Complete or remove the highlighted rules. | A rule in Query Search has no column, operator, or value | Fill in or delete the highlighted rules |
| Couldn't load the list of columns. | Query Search could not load the columns it filters on | Select Retry in the dialog's banner |
| Failed to retry all N selected fallout(s) | The bulk retry was rejected for every row | Open one fallout and retry it there to see the error |
Next steps
If the same error keeps producing fallouts, change how Engage handles it automatically: Fallout definitions for what to do, Retry configuration for how often, and Alert definitions for who hears about it.