# Automatic Error Notification for Scheduled Jobs

Treasure AI automatically emails the owner of a scheduled job when a run fails, and again when a later run succeeds. No configuration is required — the notification follows job ownership. The behavior is deliberately quiet about repetition: a job that keeps failing sends one email, not one per run, so a broken nightly job cannot bury the rest of your inbox.

This applies to both kinds of scheduled job in Treasure AI:

* **Scheduled queries** — Trino and Hive queries that run on a schedule. Notifications go to the owner of the scheduled query. See [Scheduling a Query](/products/customer-data-platform/data-workbench/queries/scheduled/scheduling-a-query).
* **Scheduled data imports** — Integration Hub sources with **Repeat** set to **On**. Notifications go to the owner of the data source. See [Data Import Scheduling](/products/customer-data-platform/integration-hub/batch/import/data-import-scheduling).


The feature offers two kinds of notification.

## Error Notification

If a scheduled job encounters an error, Treasure AI automatically notifies you by email with a message similar to the following:

```
Hi <name>, Job 123456789 has failed. It was initiated automatically at 2016-04-11 07:00:00 +0900 by your scheduled query job_title. You may manually retry and run the query as is. Alternatively, you may view and edit the query before running. We'll send you a notification when the next job for this query successfully executes.
```

If a similar error happens repeatedly, you receive a notification only for the first occurrence. A single error notification can therefore represent a long run of failures — check the job history rather than your inbox to see how long a job has been failing.

## Fixed Error Notification

Sometimes a job encounters an error situation but can recover or fix the error situation. When a failure situation occurs but is fixed without intervention, Treasure AI also notifies you by email with a message similar to the following:

```
Hi <name>, Your scheduled query job_title was failing, but Job 123456789 which was initiated automatically at 2016-04-11 06:15:00 +0900 completed successfully! While this query is fixed, you may need to manually rerun any previously failed jobs.
```

Scheduled jobs that continuously fail due to configuration errors may be disabled on the system side.

## Next Steps

* [About Notifications from Treasure AI](/support/about-notifications-from-treasure-ai) — how this notification compares with workflow, activation, and account-level notifications.
* [Viewing Job Information](/products/customer-data-platform/job-management/viewing-job-information) — check the full failure history behind a notification.