But what about servers and applications? Why spend so much time up front configuring network to make sure it doesn't fail, and then deploy an application to a single server?

Sure, there are ways to make sure individual servers have some redundancy to minimize failures -- things like RAID1, RAID5, or RAID10 (redundant array of inexpensive disks) which will protect against a disk drive failure (and I highly recommend this type of configuration for all production servers - and preferably use of hardware RAID vs. software RAID). But what happens if a file gets corrupt on RAID array? Or a recent configuration change brings application down? Or a newly released patch conflicts with other settings and causes problems? Well, in these situations server will go down and application(s) hosted on that server will be offline.

A good monitoring and alerting process will allow system administrator to detect and address these issues quickly, but still there will be some level of downtime associated with issue. And depending on type of issue, even best system administrator might not be able to immediately resolve issue - it may take time. Time during which your application is unavailable and you may be losing business due to site interruption.

So, what can you do?

A great option - and one that has recently become more affordable - is to host your application on a webfarm. A webfarm consists of two or more web servers with same configuration, and that serve up same content. There are special switches and processes involved that allow each of these servers to respond to a request to a single location. For example, say we have two servers - svr1.orcsweb.com and svr2.orcsweb.com - that have 100% same configuration and content. We could configure a special switch* to handle traffic that is sent to www.orcsweb.com and redirect traffic to either of these nodes depending on some routing logic. All clients visiting main URL (in this case www.orcsweb.com) have no idea whether this is a single server - or ten servers! The balancing between nodes is seamless and transparent.

[*note: There is also software that could handle routing process but experience and test have shown that these types of solutions are generally not as scalable, fast, or efficient as hardware switch solutions]

There are many characteristics that you should generally expect all web hosts to include:

1. High Availability: All features in world are worthless, if your site is not available when your visitors try to access it. This incorporates many different components. Do they use quality servers and hardware? Do they have multiple redundant connections to Internet? Do they have a lifecycle management program in place to ensure that you are not being hosted on a server built in 1987? (Most data centers plan for servers to have a 3 – 4 year lifecycle.)

2. Guarantee: Don’t settle for them telling you that they have best hardware and networks; ask them to back it up. Most hosts include a money-back guarantee of some sort. Given that it takes time to register or transfer domain names, develop your site, and upload it, make sure that host you choose offers you enough time to really evaluate their services. Thirty days is a pretty common guarantee. Be sure to test speed of their support responses during this period (and during different hours of day), even if you don’t need it immediately. You do not want to find out too late that host doesn’t quite live up to your expectations.

3. Support: Your host should ideally offer support 24-hours/day. Some hosts offer this via e-mail or live chat, and others offer telephone support. Keep in mind that there is usually a financial trade off for different types of support, as they each involve different costs to host, which will inevitably be passed on. If you are type of person, who needs to talk to someone on phone to resolve an issue, make sure your host offers that as a support option, and be willing to pay a little more for it. Also, as stated in number two, be sure to test your host before your guarantee expires.