We are seeing packet loss to some of our machines at the Dallas datacenter. We are investigating.

Update: We found the issue was isolated to a single machine that was the target of a large inbound flood. That machine has been null-routed to restore network stability.

We noticed poor performance on this host.  And xen was reporting errors.  We are going to upgrade the virtualization software to a more recent version.

The problem is causing downtime for some users on this host so we are doing this as an emergency, non-scheduled change.

Update 2300 UTC: Update is running now. VPSs are stopped while that completes.
Update 2345 UTC: Update is complete, VPSs are starting now.
Update 2353 UTC: All VPSs are back online, if you are still seeing an issue please open a ticket so we can check.