FDR question
Coming at this from an IT perspective, can anyone tell me what the FDR actually records?
Specifically, if it records something that was commanded to happen, would it necessarily know that the 'something' didn't actually happen, if there was no failure signal?
I'm thinking that the fuel valves apparently tried to open to allow maximum fuel throughput. Would it be possible for the FDR to record the open-valve event, but have no knowledge of the valve's failure to do so?
I realise that it's very unlikely that this would happen on both engines at the same time, but then it looks like the cause of this accident was something very unlikely.