Skip to main content
Every item ContentFlow receives becomes a run, identified by the item’s trace ID. The States page shows every run and every batch, with the logs and the document that was sent. From there you can rerun failures, one at a time or a whole batch at once.

​
Prerequisites

You need states.view to follow runs and read their logs, and states.retry to rerun them. To filter by workflow and to open a run on the workflow diagram, you also need workflows.view.

​
Run statuses

​
Find runs

Open States. The page opens on Runs from the last 24 hours. Narrow the list with:
  • the Created time range
  • Workflow and Batch
  • a pasted Trace or document ID, which searches all time
A trace ID finds that run. A Nobly Insight document ID finds every run that worked on the document: the run that created it, and later runs that updated its keywords, added a revision or deleted it, including runs that failed on it. Use it to get from a document in Nobly Insight back to its run and what the sender delivered. Runs from before ContentFlow recorded document IDs may not be found. The summary tiles count runs per status for the current filters. Click a tile to show only those runs. Auto-refresh reloads the page every 10 seconds while you watch a batch come in.
States page in the Runs view with summary tiles for All runs, Completed, Failed, Running, Retry scheduled and Pending, a list of runs with status, trace ID, workflow with version badge, batch and created time, and a details panel for a failed run showing its Overview tab with the exception message

Runs from the last 24 hours, with the details of a failed claim letter open beside the list. Its Overview tab shows the batch, the step it failed at and the error.

Each row shows the run’s status, the first characters of its trace ID, the workflow, the batch and when the run was created. The vN badge next to the workflow is the published version the run executed. A run without a badge ran an unpublished draft, for example a test run. The filters are part of the page address, so you can share a filtered view as a link.

​
Read a run

Click a run to open its details beside the list. Drag the panel’s edge to widen it. Expand to full width gives the run the whole page, and Open in a new tab opens it on a page of its own. The panel has four tabs:
Document tab titled Document sent to Nobly Insight, showing Document type Policy document from rule Policy schedules and terms, the document date from Created, keywords such as Customer number 00004400 from CustomerNo and Sensitivity Confidential from lookup row 3, the field not written, and Mapped by Policy documents, version 2

The Document tab of a completed policy schedule: what was sent to Nobly Insight, each value with its reason, and the mapping version that produced it.

When a mapping stops a document, the Document tab says The mapping stopped this document and shows the row that stopped it. Nothing was sent to Nobly Insight. View on diagram opens the workflow editor with the steps this run executed highlighted, and the version it ran.

​
Logs

The Logs tab lists one line per entry, oldest first. Errors and entries with an exception open automatically. Filter by Errors or Warnings, or search the messages. Click an entry to read it at full size, with Previous and Next to step through the run. Copy and Download export the entries in view, for a support case.
Logs tab with filter chips All, Errors and Warnings, a search box, and log lines for the Map to Insight document, Upload to Nobly Insight, Log the failure and Failed steps, with the error entry expanded

The log of a failed claim letter: the mapping produced Claim correspondence, the upload refused two keywords the document type does not have, and the exception flow logged the error and marked the run as failed.

​
Fix the cause first

Read the error against the step it happened in. Typical causes: If the fix is in the workflow or a mapping, publish it before you rerun. When you escalate, include the trace ID, batch, workflow, time and error message, and only the metadata needed to reproduce the failure.

​
Rerun failed runs

Select failed runs and choose Rerun N failed…, or use Rerun… in a run’s details. Choose how the runs start again:
Rerun dialog explaining that only runs in the Failed state will start, with the options Continue from the failed step and Full rerun, and the buttons Cancel and Start rerun

Rerunning three failed runs. Full rerun is the right choice here, because the fix was published in the mapping.

Only runs that are still Failed start. Others are skipped and listed when the rerun finishes, together with any that could not be confirmed. Keep the dialog open until it finishes; large selections are sent in parts. A rerun that started has only been scheduled. Follow the runs on the States page until they complete.
A full rerun repeats every step, including uploads, callbacks and emails that succeeded before the failure. Check what the first attempt already did before you choose it, and rerun one run before a whole batch.

​
Rerun a batch

Switch to Batches to see runs grouped by batch name, with a progress bar of completed, failed and in-progress runs. Tick batches with failures and choose Rerun failed runs…. A batch rerun includes every run in the batch that is currently failed, also those created outside the time range on screen.
States page in the Batches view with columns Batch, Runs, Progress, Failed, First run and Last activity, listing the batches Testing, (no batch), Mailroom 2026-10-02 morning and Policy export 2026-10-01 with progress bars of done, failed and in-progress runs

The Batches view. The nightly policy export has three failed runs, and the mailroom batch still has one run in progress, waiting for a scheduled retry. Guided test runs appear as the batch Testing.

​
Rerun a list of trace IDs

Rerun by trace ID in the menu takes a list of trace IDs, pasted or uploaded as a text or CSV file, for example from a support case or a log export. The page counts valid IDs, ignores duplicates, and lists lines that are not trace IDs. Rerun N failed… opens the same rerun dialog. Afterwards, the list keeps only the IDs that did not start, so you can deal with those and try again.

​
How long runs are kept

ContentFlow removes old processing data automatically. By default: Rerun failed runs well within 30 days, or resubmit the item from the sending system. The documents in Nobly Insight are not affected: they follow Nobly Insight’s own retention.

Authoring with an AI assistant

Let an assistant read and draft workflows and mappings, check them, and investigate runs, with your own ContentFlow permissions.