How to survive your first scale fail week

We thought we had our scaling sorted. We had the team, the design, the architecture and the vision. Turns out that sometimes you can't plan for some jumps in scale, and you need to react really quickly to save your butt. Does your team understand what they are seeing? Is the architecture the best it can be? Have you researched the crap out of all areas related to your graph spikes? Have we got a single magic bullet we can use? We will show you the good, the bad and the ugly of what we experienced on our scale fail week - the week we grew up very quickly.