All issues have now been resolved. Both Selfnet and Selfnet PSK are fully operational again.

The root cause was a problem with the database operated within Selfnet’s Kubernetes environment. For redundancy reasons, the database is distributed across multiple containers running on different worker nodes. An inconsistency developed between these containers, which resulted in an effectively empty database. This empty database was then imported by the central WLAN gateway, leading to an empty whitelist.

During the debugging process, the gateway was restarted. As a result, newly added IP addresses that had previously only been stored temporarily (due to strong growth over the past months) were lost.

The outage was therefore caused by a chain of multiple issues that accumulated over time and had to be identified and resolved step by step.

Actually the issue seems Kubernetes related, the gateway is not at fault.