An fiber optic interface failed last night at around
1:45 AM PST on one of our top-level switches.  Service
was restored just after 3:00 AM PST.

The failure was on a link between the top level switch
and the next level ServerIron switch that operates many
of the Kattare internal clusters.  (dns, mail, www)  We
tracked down the cause and restored service as quickly
as we could.

In the future we hope to make that link redundant, but
there are several incremental firmware improvements we
have to make first before we can do that.

Our apologies for the inconvenience!