We are aware of an issue that affected some hosted servers this evening. This appears to have been caused by a faulty network switch. The switch has been replaced and service on all affected machines is now restored.
We will monitor this for stability but we believe the issue is now resolved. We apologise for any issues caused.