Tuesday 18th June 2013: 5.58am. Just spent the last three hours patching PHP scripts to work around some bad input in a queue which was causing PHP in its VM to eat 1Gb of RAM, get OOMed, get restarted and rinse and repeat to the detriment of everything else on the cloud node. Anyway finally cleared the queue and it's propped up for another while. So tired though, just five hours of sleep to come. Note to self: you need a new server. This poor overloaded 2gb node just isn't enough to cope well with 5gb of RAM consumption between the VMs. It's going to keep tipping over and eating three hours of sleep time!