According to one of the Cisco guys, the original SE3 was tested, but did not have this AAA bug, was ready to be released. Someone managed to slip in the AAA "enhancement" in the last minute and without testing. So when it broke, it sure did took Cisco a week (or two) to pull this specific link out. According to another Cisco guy, the delay was too much "paperwork". Sad.
I kind of feel better about that. Silly as it may sound, but at least it was old fashioned human error, motived by good intentions, rather than Cisco failing to give the team enough resource. Not a good day in the office for that person though. Still, I've certainly been there. Several years ago when working at an ISP, I had loads of telnet sessions open, was rushing, and reloaded one of our customer edge 7600 routers, instead of the intended CPE router! Will never forget the viral-like spread of red alerts quickly filling up the NOC monitoring screens as a few thousand customers lost connectivity for a couple of minutes. Ouch!
