When HP put its 16GB TouchPad tablet on deep clearance for $99 this past weekend, customers came in droves. Unfortunately, the web site couldn’t handle the additional load causing embarrassment for HP on a number of levels.
HP sells monitoring tools. It also sells cloud infrastructure and cloud services. It missed the opportunity to show off both.
As Lawrence Dignan pointed on ZDNet, the problems began on Saturday, starting with shopping cart issues. “On Saturday, HP had a cart glitch that where the company’s systems wouldn’t recognize the liquidation prices,” Dignan wrote. That was only the beginning though. According to Dignan, then the web site started serving up VBScript errors. By Sunday, the site was delivering 404 errors. Ouch.
You think there might be a few lessons to be learned here? Just maybe.
First of all, HP missed a golden opportunity to show off how well its cloud computing infrastructure could keep its eCommerce system going. It failed miserably. One of the big selling points of using cloud services is that when you have an anticipated run on your servers, you can always add more virtual servers to handle the additional load.
In its new guise as a software and services company, HP hopes to sell consulting services moving forward. Presumably, they want to help companies implement cloud services that can help prevent just this type of situation in times of stress.
Secondly, HP sells monitoring tools. Presumably, using its own tools, it could have been able to track down the nature of the problem and fix it. As the ZDNet article points out, HP did eventually fix it, but was like the boy sticking fingers in the dike as a new problem cropped up for each one it fixed.
HP’s situation actually could have been a case study for monitoring tools, enabling them to show off their tools and show how well you can track down problems and fix them.
HP clearly had a bad day on Saturday. Not only did it have to unload its shiny new tablets for pennies on the dollar, it allowed server capacity to get behind demand.
But don’t necessarily get too smug about this because no matter who you are or what your company does, what happened on Saturday could happen to you too. It’s the nature of doing business on the web. You’re going to have bad days.
But if you put good monitoring tools in place that help you track the nature of the problem at the transaction level as quickly as possible, you can find the problem, and fix it and get back up and running as quickly as possible.
The point here isn’t to pick on HP. It’s to try to ensure it doesn’t happen to you — and if it does, you’re prepared to attack the problems as soon as they pop up.