At 16:34PM Friday afternoon, some subnets in Cory lost connectivity to Soda. The network recovered within 5 minutes. The border router indicated some errors; we are working closely with the vendor to find the root cause. If we need to replace any hardware to solve this issue we will announce the outage in advance.

---

UPDATE

[2008-11-21 18:06:26 | Janet Shu]

The network recovered at 16:39pm. We have made changes on the router and believe the problem should not come back.

---

UPDATE

[2008-11-25 16:35:25 | Mike Howard]

The network disruption in Cory Hall was the result of network upgrades; a loop was briefly introduced on a backbone VLAN that does not run Spanning Tree Protocol.

The error messages on the EECS border router were unrelated but are indicative of potential problems. Future downtime will be scheduled to upgrade the code on the border router to a newer version which includes a software fix which resolves the problem.

Resolved as of 2008-11-21 18:01:00