Look closely at the Microsoft Windows 2000 Server ads screaming about five nines. Find the asterisk and read the fine print. I did.
One could imagine a Titanic survivor finding out afterward that there was an asterisk after "unsinkable" and that, according to the White Star Line, you were guaranteed safety as long as the ship stayed away from all the nasty things it might encounter out in the ocean.
The Microsoft ads have taken a "shout it from the mountaintop" approach. The one I'm looking at is a spread that combines an unnatural close-up of a traffic light, green of course, with five circus-coloured nines plastered across a somewhat rickety-looking water tower.
The message is clear - near-constant uptime is within the reach of the masses. The notion of yearly downtime measured in single-digit minutes is dangled before the reader. "99.999 percent Uptime. For a server operating system . . . that translates into just over five minutes of server downtime per year.*"
But before you get too elated thinking that even YOUR server can experience such uptime, follow the asterisk to the bottom of the page.
No doubt at the behest of Microsoft's lawsuit-averse legal department, we encounter the fine print. And fine it is - about 6 or 8 point as best as I can reckon. But it is in bold type. It reads as follows: "This level of availability is dependent on many factors outside of the operating system, including other hardware and software technologies, mission-critical operational processes and professional services." What about global warming? (As I've asked on a previous occasion.)
So there it is - one very bold "uptime" statement drastically diluted by one all-encompassing caveat. Which leaves us with what? Nothing, certainly, that you can take to the bank - or to your boss.
The statement alone seems to undermine Microsoft's basic claims. One can understand that hardware (at least the basic server innards) will influence the uptime of the operating system. But isn't the Microsoft server system certification supposed to take care of that?
And, what do they mean that "other . . . software technologies" can be uptime culprits? A true mission-critical operating system doesn't let application software take control of the system in such a way that it can crash the system.
But let's not stop there. I suppose by "operational processes and professional services," they mean that if the cleaners inadvertently unplug your server to plug in the vacuum you can't hold Microsoft responsible and count that against your five minutes. Why don't they just include ". . . and acts of God?"
Naturally, if you (the network professional) trigger a shutdown, you can't count that against your uptime. I think that is the key to the whole 99.999% claim.
My own experience with Win 2000 (Professional not Server) is that where NT 4 forced you to reboot (with its blue screen), Win 2000 makes you WANT to reboot. Things just stop working - a simple drag-and-drop elicits cryptic "no more system semaphore" messages.
While, granted, Win 2000 is far more stable than its predecessor, the 99.999% claim has caveats so broad that I find it hard to take seriously.
Tolly is chairman and CEO of Tolly Research. Tolly is also founder, president and CEO of The Tolly Group. Send email to Kevin Tolly.