2026-06-03 CloudieConnect Failing to Register
Table of Contents
Event Description: At approximately 8:00 AM ET, June 3, 2026, we became aware that the IPs associated with the CloudieConnect services became unassociated with the registration servers
Event Start Time: 2026-06-03 8:30 AM EST
Event End Time: 2026-06-03 10:30 AM EST
Affected Services:
- CloudieConnect Desktop App (Branded and Unbranded)
- CloudieConnect Mobile App (Branded and Unbranded)
Event Summary
Event Timeline
June 3, 2026
- 8:00 AM EST - Tragofone reported domain unreachable, ports inaccessible, and portal unavailable
- 8:21 AM EST - Confirmed Elastic IP was unassigned from all servers
- 8:40 AM EST - Elastic IP manually re-associated to Server 2, domain briefly restored
- 8:42 AM EST - Server 2 found unreachable via SSH, Elastic IP moved to Server 1
- 8:45 AM EST - Both servers temporarily unreachable, internal IPs found to be cleared
- 9:28 AM EST - New static IP assigned to Server 2
- 9:38 AM EST - New static IP assigned to Server 1, services fully restored
- 10:30 AM EST - Root cause identified via AWS logs, automated script under Tragofone-EC2-Failover-Role performed ‘DisassociateAddress’ without corresponding ‘AssociateAddress’.
June 4, 2026
- 10:00 AM EST – After 24 hours of sustained stability confirmed, the incident was marked as resolved.
Root Cause
At approximately 8:00 AM EST, Tragofone advised that the domain was unreachable, required ports were inaccessible, and the portal could not be opened. All services on the server side were also confirmed to be running as it should. Upon investigation, it was found that the server was failing to reliably receive commands from AWS, causing the AssociateAddress command to fail.
Impact Summary
CloudieConnect Failing to Register: Approximately 2.5 hours of customer impact
CloudieConnect desktop and mobile applications were unable to register for both branded and unbranded apps. This affected all end users attempting to access the platform.
Future Preventative Action
Immediate Actions
Immediate Action that was taken was manually re-associating the Elastic IP address
Long-term Actions
As a long-term fix, we added a secondary interface for fallback and have implemented monitoring to alert if and when the primary interface becomes unreachable to prevent this from occurring again.