I'm blessed with three young sons. Anyone with young children knows that they have an unerring ability to sniff out any hint of parental hypocrisy. For that reason, I try to make sure that my words and actions reflect the principles I've tried to teach my children, and usually that practice follows over into my professional life.
Unfortunately, I got busted this week. No, I didn't break any of the Ten Commandments or download movies illegally. It was worse than that; I was rendered incommunicado in the middle of a public Webcast for MessageOne. I failed to follow one of the cardinal rules of high-availability design: Beware of single points of failure. That rule came back to bite me when I lost all Internet and phone connectivity in the middle of my presentation.
How did this happen? In a word, thrift. I get my Internet service from Buckeye CableSystem. Its service is inexpensive and fast, and the company is generally responsive to suggestions because it's locally owned and operated. About 6 months ago, the company started offering telephone service. I happily took advantage of the chance to drop SBC Communications for my business line. Things seemed to be working well, so last month I moved my residential line over to Buckeye as well. Right now, I'm saving about $70 per month compared with the cost of getting phone service from SBC.
You can probably figure out where this is going. When Buckeye had an equipment failure in my area, my Internet and phone service was knocked out at a crucial time. I was giving a presentation about improving the availability of messaging services, then I suddenly disappeared without so much as an event-log message.
Fortunately, two other principles of high-availability design helped me to avoid a total wipeout. The first principle is redundancy. Thanks to my cell phone and a Verizon Wireless data card in my laptop, I was able to call back in to the audio bridge and log back on to the presentation site in time to handle questions from the audience. This one instance more than made up for the $60 per month cost of the Verizon AirCard service. Although I had a brief interruption in connectivity, things would have been worse had I not already set up a redundant communications channel.
The second saving principle was cross-training. The Windows IT Pro Webcast staff was prepared, as was Bryan Rollins, vice president of product development at MessageOne and my co-presenter. They were able to smoothly take over, explain what had happened, and carry on with their part of the presentation. This minimized disruption to the attendees.
I learned two lessons from this experience. The first lesson is if you don't have a backup DSL circuit or other means of connecting to the Internet when primary connection fails, you should probably get one. The cost is likely minor compared with the potential benefits. Although I had a single point of failure in my communications system, I was able to recover from the failure and keep operating.
The second lesson I learned is to be sure that the people around you are able to carry on if something interrupts your usual operations. I'll have more to say about this topic next week. For now, I have to go call Buckeye and ask it to please keep up its services while I'm presenting.