Late October I did a first visit at a new client for a performance review. Not 2 hours into my onsite I just happened to be watching a server on taskman and saw every CPU jump to 100% CPU. One of the admins came to me 2 minutes later and asked if I had done anything to this server. They were getting a flood of outage messages (from their clients no less).
The outage lasted almost 12 minutes, during which time the server wasn't doing ANY work. It took me about 40 minutes but I finally isolated the problem to the one demonstrated (and fixed) by this issue. This had been heppening on a random basis for months now on several of their servers and they had been unable to figure out what was going on. The client now thinks I walk on water! I even go not one but TWO visits from the CEO that day. :-))