[Multiple Regions] - Process Mining Ingestion Failures

Update timeline

1. identified Jul 16, 2026, 10:36 AM UTC

   We are investigating an issue causing ingestion failures in Process Mining across multiple regions. The root cause has been identified, and our engineering teams are actively working on mitigation and service restoration. Customers may experience failures in data ingestion into Process Mining during this time.

2. identified Jul 16, 2026, 11:02 AM UTC

   We have applied a mitigation in the Europe, United States, Japan and Australia regions, restoring Process Mining ingestion capabilities in these locations. Also we are actively working to restore functionality in the remaining affected regions

3. resolved Jul 16, 2026, 11:07 AM UTC

   The mitigation has been successfully applied across all affected regions and Process Mining ingestion functionality has been fully restored.The issue has been resolved.

4. postmortem Jul 31, 2026, 08:28 AM UTC

   ## **Customer Impact** Between **July 16, 2026 at 10:20 UTC** and **July 16, 2026 at 11:07 UTC**, a subset of customers experienced failures when ingesting new data into **Process Mining**. Existing Process Mining data remained available and was not affected; the impact was limited to new data ingestion requests. The issue affected Process Mining data ingestion across multiple regions and was treated as a global customer-impacting event. ## **Root Cause** The incident was caused by a storage account secret rotation that interrupted access between the Process Mining ingestion services and their underlying storage accounts. Although the storage account secrets were rotated successfully, the updated credentials were not immediately applied to the dependent services. As a result, new data ingestion requests failed until the affected services were restarted and began using the updated credentials. ## **Detection** The issue was detected through automated monitoring at **10:20 UTC** on **July 16, 2026**, when an alert was triggered for the Process Mining backend service. The incident was declared customer-impacting one minute later at **10:21 UTC**. Because automated monitoring detected the issue immediately, there was effectively no gap between the start of customer impact and detection. ## **Response** Following detection, engineering confirmed that new Process Mining data ingestions were failing across multiple regions and initiated incident response. Customers were informed through the public status page while the investigation was underway. After identifying the storage secret rotation as the root cause, engineering restarted the affected Process Mining services so they could begin using the updated credentials. Service was restored region by region as the affected services recovered, and once ingestion functionality was confirmed across all affected regions, the incident was resolved. ## **Follow-Up** To help prevent similar incidents, we are implementing the following improvement: * **Automatic service restart on credential changes:** Implement a mechanism to automatically restart Process Mining service instances when critical configuration or secret changes occur, ensuring updated credentials are applied without requiring manual intervention.