Incompetence, poor planning, probably a healthy dash of managerial stupidity not listening to the engineers. And they never drill for this which is why it’s lasted so long. But having your expensive IT staff do regular disaster recovery drills is complex and expensive so almost nobody does it properly.

Recently retired from one of the major telcos. They would do DR exercises with tabletop discussions. No actual testing or hands on. Once I learned that I knew the first big disaster we had I was just gonna walk. It was that bad.