Categorized | the i.t. files

A week for disasters small and large: Google, Rackspace, and Authorize.net

Lion and Sun Ray server
There goes your Web site . . .

On July 2, Google’s app engine went kablooie for still unexplained reasons, stranding thousands of Web application startups and businesses as well as tens of thousands of their customers who use those Web apps.

Disasters, it seems, come in threes. For just last week, Rackspace, featured prominently in our book, Shoestring Venture: The Startup Bible, as the most reliable dedicated hosting service, shut down all the servers in its Dallas data center after it was hit — get this — by a truck, thus shutting down the Web sites of tens of thousands of businesses and e-businesses including major Web apps providers such as project management software service, 37Signals (also featured in our book). (In November, Rackspace went down for two hours after a car crashed into the data center — maybe they should move the data center off the median strip on the freeway.)

And just a mere 24 hours after the Google Apps engine sputtered, on July 3, the number one payment gateway in the world, Authorize.net (also recommended in our book as the best payment gateway solution for your Web site), went dark for over three hours because of a fire in its data center, which meant that literally hundreds of thousands of e-commerce sites were suddenly and inexplicably unable to process customer payments on their Web sites. While we may never know, the cost in lost business probably soared well into the millions, if not tens of millions, of dollars.

It’s enough to give an entrepreneur or small business owner agita good for three lifetimes.


What makes these disasters so stupefying is that we’re not talking about minor league players here. Google, Rackspace, and Authorize.net are best of breed in their respective Web tech fields. NetCraft, as recently as 2007, named Rackspace the “most reliable” hosting service in the world. And Authorize.net is so reliable that, well, no other payment gateway even come close. This week of Web disasters is like Tiger Woods spending an entire day driving balls into the rough, Pete Sampras overshooting all his serves, and Kobe Bryant missing every single layup — all in the same week. But these epic Web failures multiplied into a million other failures as all the Google Apps, Authorize.net, and Rackspace customers (and their customers) were. And, if your Web business sat somewhere on that failure chain, then you, too, probably lost business.

And don’t get me started on the problems Facebook has been having with outages.

So, rather than mainline a pound of Maalox, is there anything to be learned from these disasters.

It’s not just bits and bytes, it’s hard drives and processors
Marx spoke of a “fetishism of commodities,” in the sense that in a capitalist society, we are besieged with commodities but have no idea how those commodities got on the store shelves, either who built them or where. Poof! Products appear as if by magic right out of the thin air as if they were beamed down from the Starship Plenty. Be that as it may, there’s no question that we have a fetishism of the Internet. We turn on our computer and, poof, from out the thin air we have Web sites and stores and newspapers and baby videos.

However, despite the fact that it’s “digital,” the Internet is actually a physical object, or rather a whole bunch of physical objects. All data in all forms — on your hard drive, in your RAM, on your Flash drive — exists physically and is subject to the slings and arrows of misfortune. It is entirely possible, then, for a truck to drive into the Internet and shut down a portion of it. Which should be a wake-up call to you that a truck — or a toddler or a Shar Pei — could smash into your Excel and QuickBooks data, at least its physical form on a storage device, and send it hurtling to never-never land.

The “B” in “Plan B” means “Back-up”
Disaster prevention has two components, both of which are perfectly described with the word “back up.” In the first case, “back up” means creating another physical object somewhere synchronized with your data. If a truck (or a toddler or a Shar Pei or a fire) smashes into your data in one place, you have the data sitting physically somewhere else.

The second aspect of disaster prevention is also a “back up,” but in the sense of an “alternative.” If, for instance, you had set your site up to process customer payments using only Authorize.net, then on July 3, you lost business for over three hours. If, like many savvy Web managers, your site offers alternative payment gateways, such as PayPal, then you were at least able to capture some of that lost business (this is true — several sites survived the downtime by directing at least some customers to the PayPal gateway).

This, in the world of data jockeys, is called “redundancy,” and the principle works like this: if one part of the system fails, another part can take up the slack.

And so, inspired by the outages of the best data center managers in the business, part two of this post will deal with the fundamentals of disaster prevention.

Be Sociable, Share!

One Response to “A week for disasters small and large: Google, Rackspace, and Authorize.net”

Trackbacks/Pingbacks

  1. [...] outage frustrates merchants, consumers ,” Journal Star, August 4) A month ago, it was Authorize.net that went dark because of a fire in its Seattle data center. And now this. Pingdom crunched the numbers and [...]


Leave a Reply

Shoestring Book Reviews

Shoestring Venture Reviews
Richard Hooker on Jim Blasingame

Shoestring Fans and Followers


Categories

Archives

Business Book: How to Start a Business

Shoestring Book

Shoestring Venture in iTunes Store

Shoestring Venture - Steve Monas & Richard Hooker

Shoestring Kindle Version # 1 for e-Commerce, # 1 for Small Business, # 1 for Startup 99 cents

Business Book – Shoestring Venture: The Startup Bible

Shoestring Book Reviews

Shoestring Venture Reviews

Invesp landing page optimization
Powered By Invesp
Wikio - Top Blogs - Business