Runs and exceptions
Every time a trigger fires, Orderly starts a run of the flow's published version. Runs record what each step received and returned, so you can see exactly what happened. When a run fails, an exception lands in a queue where you decide what to do with it.
Open RunsThe runs list
The Runs tab lists every run across your flows, newest first. Filter by Flow and Status (and by Client, if you manage clients). Each row shows the run ID, the flow, what triggered it, its status, the client, when it started, and how long it took. Click a row to open the run.
| Status | Meaning |
|---|---|
queued | Created and about to start |
running | Executing its steps |
waiting | Paused on a Wait for event step until a signal arrives or the wait times out |
succeeded | Finished every step |
failed | A step failed and the run stopped. An exception is open for it |
cancelled | Stopped by someone before it finished |
The Flows tab sums things up at the top: active flows, runs and failures in the last 24 hours, open exceptions, and waiting runs.
The run view
Opening a run shows its status, a short run ID, what triggered it, and when. If the run failed, the error is shown in a banner.
The canvas draws the flow version the run used, with each step painted by its result. Select a step to inspect it: its Input (the configuration after references were resolved), its Output, and any Errors. Steps that did not execute in this run say so. The Timeline beside it lists every event in the run's life, from start to finish, with timestamps.
A For each step shows as one row with its counts of succeeded, failed, and not run items. Select it to page through its items, and select an item to see the steps that ran for it.
The run view refreshes on its own while the run is still queued, running, or waiting.
Waiting runs and Send signal
A run that reaches a Wait for event step pauses with the status waiting
until a matching signal arrives. Usually another system sends that signal
through the API (see Waits and signals), but you can
also send it by hand.
On a waiting run, click Send signal. The Event field is filled in with the event the run is waiting for. Add an optional Payload as a JSON object, then click Send signal. The wait step finishes with your payload as its output and the run carries on.
A signal only lands when the run is waiting and the event name matches the waiting step's event exactly. On any other run the button is disabled.
Cancel and replay
- Cancel appears while a run is queued, running, or waiting. It stops the run
and marks it
cancelled. Steps that already finished are not undone. - Replay appears once a run has finished. It starts a new run with the same
trigger payload against the flow's currently published version, then opens
the new run. The flow must be active and published. A replayed run shows
replayas its trigger.
Replay is how you rerun a payload after fixing the flow. To pick up a failed run where it stopped instead, retry its exception (below).
Exceptions
The Exceptions tab is the queue of runs that failed and need a decision.
Filter by resolution: open (the default view), retried, corrected,
ignored, or escalated. Each row shows the priority, the flow, the step that
failed, the error, how long ago it failed, and whether it is assigned.
An exception is high priority when the failing step moves money (Purchase label), and normal otherwise. When a money step fails in a scheduled flow, Orderly also pauses the flow so the schedule cannot keep failing; resume it from the flow page once the cause is fixed.
Act on an exception
- Open run jumps to the run with the failing step selected.
- Retry resumes the same run from the step that failed. Steps that already succeeded are not run again, so nothing is done twice. A failed run can be retried for up to 30 days.
- Ignore closes the exception and leaves the run failed. Add a note explaining why it is safe to ignore.
- Escalate flags the exception for the Orderly team. Describe what you need in the note.
- Correct (Mark as corrected) records a fix you applied outside the flow, with a note.
Assigning an exception to a teammate is available through the Flows API.
Retry continues the original run from its failing step, which is right when the cause was outside the flow (a carrier outage, a missing record you have since created). Replay starts over from the first step with the latest published version, which is right after you changed the flow itself.
Next steps
- Waits and signals: resume waiting runs from your own systems.
- Flows API: list runs and work exceptions programmatically.