Monitor running, failing and paused pipelines

Open Monitor and switch projects

Click Monitor in the left sidebar, from any project. The orange badge shows the count of items in Needs Attention. It appears when one or more items need attention, and disappears when all are resolved (or when you can't see any of them). It updates with each refresh.

The project selector next to the Monitor title lets you view any project you have access to. Switching here does not change the project you are working in. It is a read-only view into that project's live state.

Who can open Monitor

The four Monitor tabs

The badge colour is a secondary cue: each tab also shows its name and count.

How often Monitor refreshes

Polling returns to every 5 seconds as soon as you go back to Monitor. Items that finish between polls disappear on the next refresh. This is expected.

Act on items in Needs Attention

The tab has up to five sections: Errors, Pending, Automations, Pending Retentions and Governance. Each appears only when it has items. Error, Batch pending, Pipeline changes pending and Governance rows also have a Show in workflow icon button that opens the View's workflow. Other rows have none.

With nothing to do, the tab shows a checkmark, "All clear" and "Nothing needs attention right now."

Fix pipeline errors

Each row reads "Error running pipeline", with the dataset name, View name and task count. Click Go to View to open the View in the Data Editor and inspect the error. If the View was deleted, Monitor shows "View is missing or was deleted." See Fix a failing step.

Apply pending batches and updates

Batch pending means new data has arrived at a destination and is waiting to be applied. The row shows "Batch pending", the source View name, "New batch received 5 minutes ago", the row count and the destination or operation (for example Join: Orders 2026 or Source: Sales).

A Batch pending item appears when a Join or Lookup source View receives new data, a connector export or Publish-as-Database destination receives new data, a cross-project dependency's source updates, or a Live Link's external source changes. When the source link has Auto-sync off, the batch is held: the row reads "new batch received, auto-sync off" and shows Preview and Process instead. Items clear when the destination is updated or Auto-sync is turned back on.

Other pending items that need your input appear in the same section: Ambiguous date format, Password required (a password-protected file), Schema mismatch, Sheet selection required and Unstructured rows. Each has a button such as Select sheets, Fix batch settings or Fix settings that takes you to the item.

Update pending means a destination that isn't a dataset (for example an export) has new source data. The row shows the View name, row count and export type (for example 66,943 rows · Export: MySQL). Click Update to run it now. The row can also show Copy link to clipboard for a Live Link. After a successful update, a toast reads "Update data".

Pipeline changes pending means a pipeline has draft changes not yet applied. The row shows the dataset, View name and task count (for example 6 tasks ready to run). Pipelines with errors come first. Click Go to View to review or apply the changes.

If an update fails for some items, a toast reads "Some of the pending items could not be updated. Please verify them." Check each item on its own.

Review files an automation couldn't add

The Automations section lists files that an automation's rule sent to a dataset but that didn't go in, and that nobody has dealt with yet. A row reads File didn't go in (red icon) or File kept separate (amber icon), with the folder, file name and reason, for example "Couldn't be read" or "Didn't go into Sales 2026". Click Review to open the file in the Data Library.

The person who added the file and the automation's owner also get a notification. A file whose columns don't fit is listed under Pending instead. A duplicate isn't listed: it shows in its folder and the automation's History. See Fix files a watched folder didn't add.

Approve pending retentions

Data retention policies waiting for approval show the notice "These actions will permanently delete or suspend data once approved. Review each policy carefully before approving." Rows are labelled "pending deletion" or "pending suspension". Use Review and Approve to inspect and approve each policy.

Approve Checkpoints and review failed data checks

The Governance section appears when pipelines are paused at approval gates or data checks have failed. These items count toward the badge.

Checkpoints and data checks are tasks in the pipeline. The owner of a data check also gets a notification the first time it fails. See Pause a pipeline for approval with a Checkpoint and Remove duplicate rows and add data quality checks.

See what is running now

The Running tab lists all pipeline Views, export tasks and automations running or queued. With nothing running and no Active Retentions, it shows a checkmark, "Nothing running" and "No tasks are processing right now."

When rows are present, a banner appears at the top:

Each row shows the item name (a link to open it in the Data Editor), a status pill and detail text. Pipeline rows also have a Show in workflow icon button. Task and automation rows don't, and automation names aren't links.

Automation progress and task types

Active retentions

Below the task rows, Active Retentions lists data retention policies that are running. Each row shows the dataset name, Deleting or Suspending with the rule, and a progress bar of affected batches out of the total. A search box filters by dataset name.

Fix a task marked Stuck

Mammoth marks a task Stuck when it is unlikely to finish on its own:

1. For a pipeline or task row, click the View name to open the View.
2. Look for the This pipeline has errors banner and the failing task it points to. For an automation, open it from Data Library > Automations and read its Last run result (for a Data Consolidation, also its History).
3. Fix the cause. Common causes are expired connector credentials, a deleted source dataset, or a schema change upstream.
4. Return to Monitor to confirm the next run clears.

The Activity Log records who changed what, not run errors, so it won't show why a run is stuck.

See what has Auto-sync turned off

The Paused tab lists pipeline tasks (joins, lookups, crosstabs and sends to another dataset), exports, Mammoth Managed DB publishes and live links with Auto-sync off. Upstream changes don't flow through them, so downstream datasets may be stale. The tab notes: "Items intentionally paused via Auto-Sync controls. These are working as configured. Review them periodically to ensure nothing is left paused accidentally."

Each row shows the dataset and View name (a link), the operation type with "Auto-sync OFF", Paused since with the date, and Turn Auto-sync ON. Rows are sorted by dataset name, then View name. With nothing paused, the tab shows "Nothing paused" and "Every task is syncing automatically."

Turn Auto-sync back on

1. Check that the upstream data is still structurally compatible. Otherwise the task may error straight away.
2. Click Turn Auto-sync ON. The button is disabled and the status reads "Auto-sync on, syncing now" while the change applies.
3. The row clears once sync is active.

For the toggle itself, see Control when Views update with Auto-sync.

Activity Log tab

The Activity Log tab is a chronological record of actions in your workspace, such as creating datasets, editing pipelines, changing automations and connections, and inviting or removing users. It needs the View Activity Log workspace permission. If you don't see it, contact your Workspace Owner. See Activity log.

Monitor or the notifications bell

Monitor shows the live state of a project. The Notifications bell tells you, personally, when something you started or own finishes or fails: an export or download, a data pull, a file a watched folder didn't add, a data check that started failing. Use the bell for "tell me when it's done", and Monitor for "what needs fixing in this project". See Notifications.

Troubleshooting

Related

• Manage automations: statuses, Last run result and History.
• Activity log: filter and export the audit trail.
• Notifications: personal alerts for your own jobs.
• Control when Views update with Auto-sync: what Paused items mean.
• Fix a failing step: repair a pipeline error.