Just wanted to apologise for the outage that occurred from about 0530 BST (0430 UTC) this morning (so that would have been about midnight in NewYork, and ten PM in Los Angeles, midday in Australia I guess). Looks like we lost network connectivity for about 20 minutes from one datacentre on both links to the server in question. We're not sure yet what caused the outage, but it does appear to have been upstream of us at this point.