[2013-08-03 00:06:57 | Jon Kuroda]
At around 11:57PM, both the EECS Department websites came back with minimal staff intervention. Additionally, the web interface for the EECS mail list server, https://lists.eecs.berkeley.edu/, returned after suffering a similar fate. A small number of internally facing systems were also affected.
At this time, there is no apparent cause.
[2013-08-06 23:41:59 | Robert McNicholas]
The department websites www.eecs and www.cs and the department mailing list server lists.eecs are all hosted on a small cluster of two servers. One of those servers was experiencing problems Friday evening that was interfering with a department service. In the course of trying to resolve that problem, www.eecs was rebooted twice and was unavailable from 23:19-23:22 and again from 23:30-23:52. lists.eecs was similarly not accpeting connections from 23:37-23:56.
The underlying problem is that these systems are very old. A disk error on one node seems to have initiated the failure. We hope to have these systems migrated to a new virtual machine environment in the next few months.
Resolved as of 2013-08-02 23:57:00