Retrospective down time

[krivenbk]

I am supporting the evaluation of a continuous use system that collects data for a fixed time period then does an analysis and submits a report of the results. The developer has proposed a definition of down time for the system that forces us to retropsectively treat time when the system was up and functioning as down time if the analysis is interrupted and a result is not reported. For example, if the collection takes three hours, the analysis normally takes three hours, a downing event occurs one hour into the analysis that erases the data and it takes three hours to repair the system and bring back online, the definition being proposed would consider the down time to be seven hours rather than the three hours the system was actually down. I'm wondering if anyone has ever run into something like this and if so how did you deal with it? Thanks.