|
Internet Services
Why Choose The Internet
Factory?
When you're picking an Internet Service Provider,
you are doing more than selecting which brand of gasoline you're going to use
this week. You're actually choosing a partner in your business. When your
business has a strong tie to the Internet, it is your ISP who shows either a
good or bad face to the world. Picking the right ISP is like picking the right
employees: your ISP represents your company and you want one which is
compatible.
Unfortunately, the business of being an ISP is a fairly technical one, and it is
difficult for most people to identify the differences between ISPs---other than
price. This document is designed to break down some of the technical barriers to
help you understand what makes The Internet Factory a good partner for your
business.
To keep things short, we've broken it down into five "differences."
Keep these in mind and ask pointed questions and you'll be able to tell the good
guys from the bad guys.
1) Our business plan.
Most ISPs have a business plan which boils down
to the following: "make as much money as possible," generally adding
"and then get bought out or go public."
Our business plan is "provide the best service possible." We've been
providing consulting services since 1995, and Internet services since 1998. This
plan seems to be working.
So what's the difference? The difference is that quality service is the most
important thing in all of our staffing, buying, and technology decisions. Not
cost. Not saving money. But service.
That means that our staff are all high-level network consultants. There's no one
working for The Internet Factory who is a "technical support person"
we hired for $15/hour. When you call us and talk to us, or when someone here
acts on a support request, you're dealing with competent, professional, and
dedicated people.
We work very often with other ISPs and are amazed at the quality of people we
have to interact with. Generally, an ISP has a tiny core of very few, very
bright people (although we have seen ISPs who have no bright people at all) and
a large ring of people who have absolutely no clue how the Internet works. The
goal of the ring is to keep you from talking to the core, and this means that
getting a good technical answer or quick results out of most ISPs is a long-term
exercise in frustration.
We don't do that. When you get someone in email or on the phone, you're talking
with our best and brightest. It's more expensive to do things that way, but
that's the way we want to do things.
2) Our network engineering plan.
We don't over commit ourselves. If you do a quick
analysis of the way an ISP works, you can see that there's a very, very strong
fiscal incentive to cram as many customers as you can onto your network. That's
why we laugh when we see ads for huge hosting companies bragging that they have
18,000 (or more!) domains hosted. Of course they want more business: it's all
profit!
We don't work that way. We have a fairly luxurious set of resources: 10 Mbits of
external bandwidth, a moderate number of digital dial-in lines, over 100 Gbytes
of disk storage, and a completely redundant network. That costs a lot of money
to buy and maintain, but we do it because it gives us a tremendous tool to bring
to bear on the Internet connectivity our customers want from us. Yet we only
have about 300 customers.
Our customers have to shoulder a larger cost burden than those paying $20.00 a
month or more, but they get much greater resources for their dollar. In fact, we are
cheaper than the biggest of the big---you get a lot more bang for your buck.
If you're running a laundromat and want a single page to tell people that washes
are still $0.75, you probably don't need us and you probably don't want to pay
for us. But if you want to make a credible presence on the Internet, communicate
quickly, reliably, and effectively with your customers, and have the option to
grow quickly, well, we're here for you.
3) Our disaster recovery plan.
We have one. The people who own and run
The Internet Factory aren't kids just out of high school who think it's
cool to upgrade their Linux servers once a week. We're serious data
processing professionals who come from an environment where the simplest
applications are considered mission and business critical and the most
important ones have to have FDA approval because someone's life is on
the line.
The Internet traditionally has taken a loose view of reliability.
Started as a research experiment, the ethos of the Internet has been
that if it's not up now, it will be later, and that's O.K. Well, that's
not O.K. with us. Several ISPs have been advertising "99%"
uptime. That figure is laughable. We could be down for two entire
business weeks, all day, every day, and still be up 99% of the time.
In our computer room, we have designed so that no single failure will
stop us.
* Uninterruptible power systems are ready to take the load for up to an
hour. An off-site generator can keep us going as long as we can put
diesel in it.
* Our systems are constantly monitored by automated test systems. Should
any system fail to respond, an automatic series of escalation procedures
occur, starting with paging people and working up from there.
* Every computing element is doubled. Using high-end clustering
technology, we know that if any CPU, disk, controller, network card,
hub, router, or bridge were to fail, things would automatically keep on
going. And not with a "failover" interval---it's always hot,
always clustered, always non-stop.
4) Our security.
We believe in it. Have you ever wondered what's
at the other end of that so-called "secure server" you're sending your
credit card number to? Well, it's probably a Unix system. Unix, powerful,
speedy, inexpensive, cryptic, and widely available. Also widely known to have
poorly designed and implemented security. Why bother to capture credit card
numbers as they fly over the Internet? Just break into the host they're stored
on and you get them by the pile.
CERT, the organization which issues Internet-wide reports of serious security
bugs, has issued 106 Unix alerts. During the same time, there were 4 OpenVMS alerts.
The Internet Factory uses OpenVMS. Now you know one of the reasons why.
We have had many security attacks---none of them successful---in the years we've
been in the Internet business, and we've been hired by companies to clean up the
mess after someone else broke into their networks.
We don't mean to suggest that ISPs running Unix are insecure. There are lots of
high-end ISPs who understand how to build and manage secure systems. What we're
pointing out is that we have systems which are proven, over time, to have fewer
undiscovered security bugs in them. Even the best written application is a
security problem if the operating system it's built on has a hidden
problem---which is discovered by the wrong people.
Our seriousness about security extends to our clients as well. We carefully
evaluate customers and their applications. Will this draw unnecessary attention
to us? Could this be the subject of an attack? The Internet Factory, as a
company, has nothing against adult web sites, radical political viewpoints, or
even hackers. We just don't want them to be in the same room with our systems.
There are lots of ISPs who will enjoy taking on that risk---we're not one of
them.
Similarly, we don't invite users to build their own web applications. It's too
easy for a simple mistake to lock up an entire server and affect everybody using
that system. If you feel that your business cannot survive without you running
your own Perl scripts on your web server, we'd be happy to help you find a more
appropriate hosting provider.
5) Our marketing plan.
We depend on word-of-mouth to bring businesses to
us. This has two benefits: first, we keep the flow of new business to a
manageable level. We don't want to suddenly have 50 new customers---we can't
give good service that way.
Second, we know that when someone calls us, they already have the best
advertising we could possibly have: a trusted recommendation. This works both
ways: we don't want to work with clients who don't fit with our business model
and who don't share the same goals and priorities.
|
