Peter's Griddle

Computers are terribly reliable, in general. Today's computers execute millions of instructions each second, with an error rate that is inconceivable in other technologies. Yet, if you have hundreds of thousands of machines, you do need to take care of failures.

A Cnet article elaborates on the Google situation (a Google cluster has several thousands of machines):

In each cluster's first year, it's typical that 1,000 individual machine failures will occur; thousands of hard drive failures will occur; one power distribution unit will fail, bringing down 500 to 1,000 machines for about 6 hours; 20 racks will fail, each time causing 40 to 80 machines to vanish from the network; 5 racks will "go wonky," with half their network packets missing in action; and the cluster will have to be rewired once, affecting 5 percent of the machines at any given moment over a 2-day span, Dean said. And there's about a 50 percent chance that the cluster will overheat, taking down most of the servers in less than 5 minutes and taking 1 to 2 days to recover.

These look like interesting planning assumptions for both hardware and software planners. As they say:

"Our view is it's better to have twice as much hardware that's not as reliable than half as much that's more reliable," Dean said. "You have to provide reliability on a software level. If you're running 10,000 machines, something is going to die every day."

At the Westminster eForum on Web 2.0 last week in London, Jim Cicconi, chief lobbyist at AT&T warned that the Internet will be fully clogged by 2010.

When I worked at AT&T Bell Labs around 20 years ago, the phrase "imminent death of the net predicted" was already a running joke, so something else must be going on.

In the past 20 years network bandwidth has grown by something like a factor of 1 million, and there is no end in sight.

The real reason for his statement is probably to warm the public to volume based pricing. Telecom companies hate the current flat-rate Internet pricing model because it limits their growth. Once you have served all households in your geographical area your revenue is capped, and you can only compete on price. With a volume-based pricing model you can hope to seduce your customers into buying more, and more expensive services. However, the ghost of flat-rate pricing is out of the bottle, and customers will not like going back.

If anything, the telcos should realise they are in two adjacent businesses. They supply high-volume information technology services, such as phones and desktops that work. They are also in the content distribution business. They should buy a company like Akamai, so that they can optimise both the network traffic and the user experience.

Websites can go down. But there is a lot more that can go wrong with all your digital assets online. Have you ever heard about site-defamations, spoofing, identity theft, plagiarism, and software vulnerabilies?

How much revenue will you lose, or damage will you suffer, if any of these happen? If so, do you know how to protect your assets against these risks, without paying an arm and a leg?

I am developing an e-book that spills the beans on the risks that your website, brands, and online identities run, and the sometimes surprisingly easy ways in which you can protect yourself.

This is your chance to get your questions answered. Just ask your most burning question and you will receive the e-book for free when it is ready.
Click here to claim your copy.

p.s. there is no obligation. The e-book is a reward for people who give me their questions.