Great track record

In the last 10 scheduled downtimes, SEVEN ended over 60 minutes late.

If it's taking you over an hour to fix a single switch (and I don't care how big it is---entire switches do not break, separate switch components do), then it's absolutely clear that one or more of the following three things is true:

1. you do not have extra components around, ready to swap in case of breakage
2. you're lying about the cause of the problem
3. you're actually a bunch of ferrets, randomly chewing on cables until it works again

Unacceptable. Pay your project managers more--and more importantly, LISTEN TO WHAT THEY SAY.

Someone's obviously never dealt with any kind of technical equipment. One measly hour is nothing. In fact to me, that shows they are very much on-the-ball and putting in real effort. I know Linden Labs will never do this, but I wish they would publish some pictures of the racks upon racks of severs and network switches which are the core of SL itself. Then maybe some of these people would realize just what a job it is. It's not just a desktop PC with a reset switch on the front.

There's only been one downtime in recent memory that ran 'long'. As I recall the grid was down an extra 6 hours that day. Even then, it was understandable as they had to get one of their vendors involved, I still consider in that case they did well. Most vendors are good to get back with you the next day if ever.