One day, I came into work to find a Blue Screen of Death. This wasn't good. Undaunted, I rebooted the system. When the system rebooted and ran just fine, I contented myself by checking our backup status. All was good. I ordered a new drive and was prepared to do an upgrade.

The next morning, I came in to find the blue screen again. This time, the system would not reboot. It was, to use a famous technical term, hosed.

Welcome to the ninth circle of Hell
We've all had to deal with server crashes. It's no big thing. Yes, it'll take time and possibly ruin your whole day, but if you've been disciplined, the server will come back to life. Replace the parts, reinstall Windows, run the updates, run the patches, install the drivers, install the apps, restore the backups, have some chocolate, and you're back.

That didn't happen this time.

Sure, I popped out the bad drive and put the new one in. I did all the aforementioned reinstall stuff and got a fully working Windows 2000 Server back online.

Then I attempted to restore the Exchange backup. In fact, for the next thirteen days, I attempted to restore the Exchange backup. I'd work all day, grab a few hours sleep, eat at my desk, and image, reinstall, update, and try again, over and over again.

I followed all the documented steps. Then I followed a whole lot of undocumented steps. I spoke to the very best Exchange experts in the world (that's one benefit of running a magazine on Outlook and Exchange) and each of them, in their own words, told me how they shared my pain. All chuckled a bit uncomfortably and wished me luck.

To a person, they all told me that Exchange is great, except when you have to restore it. Each told me that the best policy is to just hope you never have an Exchange failure.

If you listen quietly, you can hear Dante laughing.

The stakes
Here at ZATZ, we live on email. Fortunately, the journals themselves are mailed out using a different technology, so the journals could, in theory, go out even if the Exchange server was down.

But virtually all of our company's business is transacted through email. Virtually all of our company's institutional memory is contained in our mail store. Phone numbers, contacts, complete transcripts of deals are contained in that mail store. We had active business that wasn't getting answered, replied to, or completed because the mail couldn't get through.

Although I threw up a temporary mail server a day into the process, if I couldn't get the mail store back, our entire business was at risk. All because we put our reliance on what was supposedly an enterprise-class best-of-show product.

For thirteen days, we didn't know whether we'd make it. This, in fact, was why I stuck with it all that time. Most people I spoke to who'd had similar problems said they simply tossed out the bad systems, took their losses, and moved on. I wasn't going to do that. I was going to recover our business data and we were going to make it out of Hell.

After a while, I was convinced I was losing my mind. I wondered around, nearly in tears, saying "But I did it all right. I did everything I was supposed to. It's not supposed to be like this."