Points Of Failure

Back in 1988, I was a Broadcast Producer with what was then called the Armed Forces Radio & Television Service. I was stationed on a small island in Greece you may have heard of...called Crete. Basically, Crete is the closest thing to Eden in real life. Unless it's winter. In the winter, all of the tourists flew home, the clubs closed, the temperature dropped and we would basically invent ways to amuse ourselves as the winter months passed. Hence, the idea popped into my head "Hey...I should rebuild the engine in my car"!

There was an auto hobby shop on my base, complete with all of the tools someone who knew what they were doing could use...to rebuild or repair just about anything. I had a Chrysler "K" car, arguably the most unassuming automobile ever built. In the summer of '88, I set out on a quest to turn this mild mannered auto...into a muscle car! (There really IS a tie in with your video streaming strategy here...but stay with me...we'll get there).

Every evening, after leaving the television station...I walked down the road to the hobby shop. I greeted Scotty, the manager...donned my overalls, grabbed my government issue tool box and went to work. The next few months were fraught with bloodied knuckles. aching muscles and I even invented a new expletive or two. I pulled the engine, steam cleaned it...disassembled it and made a few "adjustments".

Now every self-respecting hot rodder knows that there is a limit to how much performance improvement you can get out of what they call "bolt on" enhancements. You know...things like fancy headers, and shiny air cleaners and anodized and braided hoses. These are cosmetic improvements to the true muscle car aficionado. If you want to really make a car faster...you have to take more drastic action. First, you mill the head! By reducing the amount of space each cylinder head has...well, you take the first step to increasing compression...which means a more powerful explosion inside the cylinder with every revolution of the crankshaft...which ultimately means...MORE HORSEPOWER! Oh, but I couldn't stop there. My next move was to lope the valves and port the cylinders (or is it port the valves and lope the cylinders?)...then there were new cylinder rings and a myriad of "added value" steps I took...all in an effort to dramatically increase the amount of power I could get out of this decidedly un-powerful auto.

One early Saturday morning, after the engine had been returned and the mess I made of the garage cleaned up...I took my place behind the steering wheel, inserted the key...and with an approving glance and crossed fingers from Scotty...turned the key. It sounded different. It sounded POWERFUL! A quick compression test confirmed that yep, I had indeed...dramatically increased the compression. I had succeeded in turning a "K" car...into an "OK...wanna race?" car! Or, had I? I rolled slowly up to the main street of Gournes...(just south of Iraklion and just north of Hersonisos for those locals tuning in). I looked left, looked right...and cautiously entered the main street. After one last glance to make sure there were no police or ox carts in the vicinity...I applied rapid pressure to the accelerator. I floored it man! Whew! With a plume of blue/white smoke and a squealing that never seemed to end...my K car dug in...and seemed to wrinkle the pavement with an explosive burst of front wheel drive mayhem. I tore down the street...quickly passing 40, then 50, then 60, then 80, then 100 miles an hour. I was just producing a grin of glee thinking of how odd this must look to the local Cretans...when I heard some very confusing noises coming from under the car. There were odd vibrations as well. There was also smoke produced by something other than my tires. It all got worse as I limped back to the shop. After hearing my story, Scotty's hand rested on my shoulder and his sage words doused my fire: "Yep, you made the top end so good...you ruined your bottom end. That drive train ain't made to handle so much torque. You're, well...screwed".

And so it is with streaming my friends. Many prospects I speak with each week are looking to do exactly what I did with my "K" car. They want more power. And they put all of it up front. They invest in an application...or install an encoder and open a live stream...all with no consideration of the weak links in the chain. If you're putting an encoder in your facility, and you want to push out three different streams...each with a bit rate higher than 500K...well guess what...your "bottom end" better be able to handle more than 1.5 Megabits of throughput. (In this case, your bottom end is your amount of dedicated bandwidth available to the outbound signal from the encoder). If you're hoping to handle multiple instances of Flash Media Encoder, each running high bit rate video streams...well then, you need to beef up the processing power on the PC you're running the encoder on. These are just a few, but there are many...weak links. If not considered when you fashion your streaming strategy...they can...well, make odd noises and generate smoke...and ultimately cause you to fail.

So, if you want to build a streaming solution that has lots of horsepower...you have to be wiser than I was in 1988. You have to consider how one piece of your system might negatively impact another piece...BEFORE you step on the accelerator.

I like to imagine that somewhere in the hills of Gournes...on the northern shore of the Island of Crete today...there is a grandfather telling a story about the morning 21 years ago when he heard a horrendous screech, looked out the window...and saw a crazy American flying down the main road in an odd looking car. Or, perhaps he would just webcast his story...to his extended family in the U.S., Japan and Australia. :)