Previous incidents

We are currently investigating a service disruption impacting a subset of users in certain regions. The issue is caused by an Anycast routing anomaly where traffic is being directed to an unintended edge location due to inconsistent BGP route propagation from one of our upstream providers.

As a result, some requests may experience increased latency, intermittent failures, or connection timeouts. Our systems are functioning normally, but traffic routing behavior is not optimal in all regions....

We have fixed the BGP Session issue and finalized our Network Maintenance

The hardware failure affecting our Tokyo POP has been fully resolved. Dashboard access has returned to normal for all users.