On April 8, 2025, at 18:14 UTC, routine maintenance at the Miami (MIA3) data center, specifically involving network ingress hosts, led to a significant disruption to both inbound and outbound network traffic. The immediate impact was mitigated by 19:45 UTC through the disabling of the affected ingress hosts.
This maintenance was not expected to cause any impact to production traffic. Initial pre-maintenance testing and the early phases of the maintenance procedure were completed successfully, without any indication of risk or instability. However, some investigation has since pointed to misrouting and overloading of specific ingress hosts as a contributing factor to the disruption. A more thorough root cause analysis is ongoing to fully understand the underlying conditions and contributing factors.
We are actively investigating the behavior of the impacted networking infrastructure and working to reproduce the issue in a controlled development environment. This will help us identify the root cause and refine our maintenance procedures to prevent similar incidents in the future.
Additionally, we are expanding ingress capacity at MIA3 and planning the deployment of additional spare ingress hosts to support future growth and improve resilience.
This summary provides an overview of our current understanding of the incident given the information available. Our investigation is ongoing and any information herein is subject to change.