The Industry Standard - Communications: Why do we accept less than 99.999%? - Comments

Interesting article.

Lawrence — Sun, 09 Nov 2008 13:42:32 -0800

Interesting article. Internet TV and PC satellite TV are the future of TV.
They allow one to download software to one's computer or laptop and watch TV online - up to 5000 channels in high definition. Find out more about it at http://www.1-satellite-tv-facts.com

That's funny, because my

Thea — Thu, 24 Jul 2008 16:52:23 -0700

That's funny, because my companies 99.99 uptime- my 8 hours of downtime a year- must be happening all at the same time. And at the wrong time! 5 hours of being down consecutively can cost someone valuable business.

Guys, what is TIO?

David V Ramos — Thu, 29 May 2008 13:58:38 -0700

Guys, what is TIO?

AFAIK the answer to the "Did

Stefanie P Roberts — Thu, 29 May 2008 13:57:50 -0700

AFAIK the answer to the "Did you install Drupal's spam module ?" is yes

Guys, what is TIO?

David V Ramos — Thu, 29 May 2008 13:57:47 -0700

Guys, what is TIO?

Larry has posted a follow-up

Ian Lamont929744 — Tue, 11 Mar 2008 08:49:22 -0700

Larry has posted a follow-up piece entitled "Communications: Is price all that matters?". It addresses many of the points brought up in the reader comments above, and on Slashdot and Paul Mah's TechRepublic blog.

Hi Larry, It's funny; I've

Samuel Pedigo — Mon, 10 Mar 2008 19:44:26 -0700

Hi Larry,

It's funny; I've been wondering the exact same thing for a little under a year now. I'm a student at a branch university of Texas A&M, majoring in computer science and minoring in business. I also plan to pursue a Master's in Business Administration. Under the honors program we're to do what's called the "Project of Excellence", and I've been doing a research paper combining both business and computer science aspects.

Did you know that only one in three software projects are completely successful? Over one in two projects are either over-budget, late, or are missing features. Twenty percent of projects aren't completed at all. I took a look at your industry experience, judging from that, you should be aware of the problems. For the other readers out there, think about how you have to wait an extra two years every time a Halo game is "supposed to come out". In my preliminary paper I focused on "The Consumer Effect on Software Development". Here I pondered a similar theme that you presented here in your article:

"Through a combination of consumer apathy and negligence, underachieving methods and the ensuing lack-luster products remain in place. Industry should and cannot be held accountable for a lack of development. As previously mentioned, the burden is on the consumer. The purpose of every profit-driven corporation is to maximize stockholder wealth. In a consumer-driven market the only way to achieve this goal is to please the customer."

If customers demand for a higher quality product, then that's what they'll receive. Sure, I have a few suggestions for members of industry, but a large burden of the responsibility is placed on us. Even still, we have to be prepared for the consequences of what we ask for. An increase in quality probably leads to an increase in price, too.

Thanks for reading,
Samuel

Larry Borsato708924 — Mon, 03 Mar 2008 19:24:31 -0800

The 99.999% is just a number, and I don't expect that companies will suddenly start providing it. However, I was just wondering why we - customers as a whole - don't demand more than we are getting. ISPs block traffic and we sit back and take it. When we start to use the bandwidth that we've always been promised and paid for, we are told we can't have it.

We should be demanding a guaranteed and audited reliability for these services.

The answer to the question

Roger Wolff — Mon, 03 Mar 2008 07:05:21 -0800

The answer to the question of why we don't see 99.999% uptime in Internet and Mobile phone markets is simple.

First, if some service happens to have 99.9% uptime (a factor of 100 too much downtime). this amounts to about 8 hours a year. The average consumer has a very small chance of actually noticing this downtime at all. Even if the user notices, it is shrugged off, as "will probably not happen again" and it's likely that other users will notice the next outage.

Secondly there is the "cost" in transferring to another service. Loss of phone number or email address, getting used to the different type of billing, arranging the transfer etc etc.

Then because outages are relatively rare, nobody has the vaguest idea about who has the best record. So changing might actually mean you're worse off. So people are again reluctant to switch.

And if someone switches, it's just one client. Because so many others end up not changing, the providers will have to differentiate themselves on other items than reliability to get marketshare.

I work for a web hosting

Chris Wik — Mon, 03 Mar 2008 04:11:52 -0800

I work for a web hosting company who offers both low-end bulk hosting and high availability hosting options.

Our costs for mostly reliable, mostly fast bulk hosting packages come in at under £1 per month per customer. We host thousands of customers on a single machine and carefully monitor resources to ensure we're always running at near-full capacity.

Our costs for offering "five-nines" availability with no single point of failure, off-site backups etc. come in at over £1000 per month at an absolute minimum. The equipment used to offer this level of service is usually 90% idle, but that extra capacity is required to ensure we can handle simultaneous equipment failures and unexpected spikes in traffic.

There is a very simple reason why services aimed at the average consumer do not offer 99.999% uptime guarantees -- it is simply too expensive, and the consumer will not pay for it. How many businesses or government offices rely on mobile phones or cable TV? Until they do, you won't see 99.999% uptime.

Speaking as a consultant

Ben - — Sun, 02 Mar 2008 14:16:08 -0800

Speaking as a consultant who specialises in development of embedded systems, it's not about what individual consumers demand - it's all about the cost of system development, or more precisely, the amount the service provider is willing to spend in prevailing market conditions. With embedded, safety-critical hardware, if the system fails to operate predictably, consistently and in a fault-free manner, people can be seriously injured or even killed. The possibility of drastic consequences of software failure leads to better quality assurance, and external hardware verification to ensure that software is operating correctly.

All of this extra development effort, as with everything in life, is an additional cost to the company, and they can do one of two things - fail to ensure software quality, and swallow the profit; or be forced to compete on quality grounds, be it from legislative or market requirements, and suffer diminished profits, or charge extra. Most software vendors tend to do the latter, and you can be assured that service providers will not spend a single cent more unless they're guaranteed to make substantial return from what could be seen as a high-risk activity.

Mobile phone and pay-TV service providers would likely not see a return from promoting "99.999% reliable service!" to the everyday consumer - as mentioned in the article, consumers just don't care, and more to the point, will not pay substantially more for a service that is available for that extra 0.099%. Such advances in reliability will only occur if consumers as a whole start demanding better quality service, or the industry unilaterally decides to pursue a strategy of competing on software quality, not price.

Communications: Why do we accept less than 99.999%?

Larry Borsato — Thu, 28 Feb 2008 10:14:45 -0800

Pick up your phone. Go ahead and try it right now.

I'll wait.

Since you're still reading, I know that you have a broadband internet connection, as opposed to a dial-up ISP. I also know that when you picked up your phone you heard a dial tone. That's because telephone service is regulated. Operators are expected to achieve "five nines" of reliability or "uptime" -- the service must be available 99.999% of the time -- and they must report any instances of downtime longer than 2 minutes. That's a miniscule five minutes of downtime in the 525,600 minutes in a year.

If that doesn't sound like much -- if you think that 99.0% ought to be just fine -- then keep in mind that 99.0% availability means that your service might be down for more than three and a half days every year. Calling 911 doesn't seem so helpful in that situation, does it?

We've come to expect, and to depend on, that level of service for our home phones. That's because local and long distance phone lines are regulated.

So why are we so unconcerned about it for other forms of communication?

ISPs used to promise unlimited internet service, but when we actually tried to take them up on their promises they complained that there was a bottleneck; they would have to filter. Or they would charge us higher prices for more bandwidth. And we seem willing to accept such explanations.

We're so used to cable and satellite television reception problems that we don't even notice them anymore. We know that many of our emails never reach their destination. Mobile phone companies compare who has the fewest dropped calls (after decades of mobile phones, why do we even still have dropped calls?) And the ubiquitous BlackBerry, which is a mission-critical device for millions, has experienced mass outages several times this month. All of these services are unregulated, which means there are no demands on reliability, other than what the marketplace demands.

Why don't we demand more?

Recently, when attempts were made to protect net neutrality, telecom companies complained that people were trying to introduce new regulations. Telecoms don't like regulations; they cost money.

Reliability requires redundancy (at least two of everything) and failover plans so everything continues to work even when something breaks. BlackBerry service went down because Research In Motion routes all messages through a single point of failure. And it failed. Systems today are built as cheaply as possible, and it shows.

We're building products and services as quickly as possible, as minimally as possible. Everything is designed to maximize profit. And, one could say, designed to fail. We're building systems to which we are increasingly entrusting our lives. Even failing only 1% of the time just isn't good enough. Would it be acceptable if the wheels feel off of 1 out of every 100 cars?

We once understood the value of just enough regulation to ensure the secure operation of services that we depend on everyday. Now we are building entirely new classes of products to replace the old ones -- services we have come to depend on even more in our everyday lives. Could you live without your mobile phone? How would you function at work without email access?

What does it take for us to recognize that we need to take the same steps necessary to ensure that the technologies created today -- and into the future -- are just as dependable as the good old telephone was 50 years ago?

Related news, commentary, and predictions:

Follow-up by Larry Borsato: Communications: Is price all that matters?
Prediction: AT&T will win the C-block spectrum auction
Larry Borsato: The wisdom of clouds
Larry Borsato: ISPs, the music industry, and the Web
Larry Borsato: If you want a job, we don't want you
News: Sprint said to be planning layoffs
News: Sprint to open its XOHM WiMAX, still faces hurdles
News: T-Mobile, AT&T join Verizon on unlimited plans

Note: Anonymous comments on The Industry Standard are disabled. To leave a comment and participate in the Standard's prediction market, please register first.