Operator notifications
Operator notifications email your team when privacy request processing needs attention. Configure recipients in the Admin UI or through the API.
Configure recipients in the Admin UI
- Open Privacy Requests.
- Click the bell icon (Configure alerts) to open the Request notifications drawer.
- Turn on Enable email notifications and add the Email addresses that should receive notifications.
- Under Notification frequency, set how many privacy request errors trigger the error notification.
- Click Save. A "Notification settings saved successfully" confirmation appears.
To stop all operator notifications, turn off Enable email notifications and click Save. This clears the recipient list.
Configure recipients through the API
Send the recipients and the error threshold to the privacy request notification endpoint. This requires an access token with the privacy-request-notifications:create_or_update scope.
curl -X PUT "https://<your-astralis-host>/api/v1/privacy-request/notification" \
-H "Authorization: Bearer $TOKEN" \
-H "Content-Type: application/json" \
-d '{
"email_addresses": ["privacy-ops@example.com", "dpo@example.com"],
"notify_after_failures": 5
}'The response echoes the saved configuration:
{
"email_addresses": ["privacy-ops@example.com", "dpo@example.com"],
"notify_after_failures": 5
}email_addresses: every operator notification is sent to each address in this list. Send an empty list to stop all operator notifications.notify_after_failures: how many unsent privacy request errors must accumulate before the error notification is sent. It applies to that notification only.
GET /api/v1/privacy-request/notification returns the current configuration and requires the privacy-request-notifications:read scope.
Notification types
The wording of these emails is fixed. They are not part of the email template types that can be customized in the Admin UI, and property-specific messaging does not apply to them.
| Notification | Trigger | Repeat behavior |
|---|---|---|
| Privacy request error notification | Unreported privacy request errors reach the Notification frequency threshold | One email per batch of errors. Reported errors are not counted again. |
| Privacy request completed with skipped erasures | A privacy request completes, but erasure was skipped for one or more datasets, typically because their integration was disabled or deleted mid-request | One email per request, naming the skipped datasets |
| DROP alert (enterprise) | A step in California Privacy DROP (Delete Request and Opt-out Platform) processing fails | Most persistent failures trigger another alert on each scheduled run, and a single run can send several alerts. Deadline alerts report newly overdue erasures and do not repeat for the same erasures. |
Troubleshoot DROP alerts
Where DROP supplies a remediation hint, such as generating a new key from the Data Broker Portal, the alert includes it. For failures that originate in your warehouse or in Astralis itself, the alert names only the kind of error; the table shows where to find additional error details.
| Problem | Where to investigate | Next action |
|---|---|---|
| Download fails with an invalid API key | DROP cycles tab, the failed cycle's details | Generate a new key in the Data Broker Portal and update the connector |
| Download fails with access denied | DROP cycles tab, the failed cycle's details | Confirm payment is complete in the Data Broker Portal |
| Download refused because the previous batch is incomplete | DROP cycles tab, earlier cycles still awaiting completion | Resolve rejected uploads or held erasures on those cycles; the next run uploads the outstanding responses |
| A response upload is rejected | DROP cycles tab, the cycle's file upload attempts | Correct the file, then use Amend on the cycle |
| Held erasures pass the 45-day deadline | DROP cycles tab, the cycle's Pending erasures count | Check why the linked erasure privacy requests have not completed |
| The match-index backfill fails | Mapping tab, Backfill status (error text is truncated) | Fix the gather query or the Snowflake connection, then run the backfill again |
| The connector is misconfigured (invalid secrets, no backfill configuration, match index not built) | DROP cycles tab, the skipped cycle's details, and the Audit log tab | Fix the connector settings, or run the backfill to build the match index |
| The worker processing a cycle crashes | DROP cycles tab, the failed cycle's details | Amend the cycle to rerun it; contact support if it keeps failing |