At approximately 2:20 a.m. Friday, Jan. 4, one of the University of Minnesota's national network infrastructure peers, National Lambda Rail, inadvertently initiated an overload of messages to its network affiliates while performing planned maintenance. This flood of additional messages sent the University's network routing table (a database used to calculate the destinations of incoming messages), over capacity, causing it to stop processing traffic.
Network engineers recovered the network starting around 4 a.m., and worked to adjust the route table capacity. Full network recovery was completed by 5 a.m.
National Lambda Rail is used to by the University, as well as several other research institutions nationally, to provide high-speed, wide-area Internet access to its network. Several other national research institutions also were affected by today's event.
All IT services are functioning normally with no long-term or permanent effects. University network analysts are evaluating further safeguards to mitigate this issue in the future. A more detailed summary of this event will be provided in the next couple of days.