In the past six months, all three sites went down – at the wrong time – because their databases couldn’t keep up with consumer demand. In every case, unplanned peaks in user traffic pounded the site, and brought performance to a screeching halt.
In this new world of immediate experience and gratification, the database is rapidly becoming the company. Consider Amazon and Pinterest: There’s no storefront, no customer service – they live and die based on each user’s experience. More than ever, that experience is backed every step of the way by data, which means that the dark corners of data performance risk are becoming a bet-the-company vulnerability.
If your data isn’t killing your database today, it will. Maybe next month; maybe next year.
To add a little perspective, one of the largest wireless providers in the UK – O2 – is currently spending close to $20 million in an effort to replace its database following two recent outages that impacted millions of customers. Derek McManus, COO of O2 UK wrote in a blog post that the company is, “…extremely disappointed to have let our customers down again.” He adds that O2 UK is “…not prepared to risk this happening to our customers for a third time and are implementing a proven alternative solution.”
On the flip side, you can get ahead of this chase for big data. It may mean rethinking how big databases are built for growth, because the underlying assumptions of predictable behavior are obsolete.
For just about every company, managing and exploiting the potential of ‘big data’ will be the #1 business opportunity before 2013 is out. Even if your database doesn’t seem like an under-performing asset today, it could easily become one within two or three quarters.
The biggest issue is performance – and staying ahead of the tidal wave of transactions. That’s especially challenging when the ‘conventional wisdom’ about big data points to the importance of analysis. Indeed, it’s the analysis that can be mined for sentiment, patterns, and, once you can decode it, predictability, new opportunities and money.
But it’s the transactions that underlie this intelligence. And it was the breathtaking volume of transactions that caused Amazon, Pinterest and Apple to go dark. Simply too much unexpected activity, likes, posts and uploads. And that – like it or not – is the future of big data. Peaks. Valleys. And not always when you expect them.
Avoiding this mineshaft is, surprisingly, possible. It’s not expensive, either. But it does mean toppling a few sacred cows.
For example, companies like Mozilla – which recently launched an online app store – are putting database scalability and performance systems and solutions in place before they build or launch new apps – in order to guarantee the company will be able to support its anticipated transaction volumes.
Size matters. But not the way you think.
To read the rest of the article, click here.