Internet Application Delivery

October 15th, 2010

Before continuing on with the ‘Cloud’, I think it might be worthwhile to take a step back and consider the history and some of the alternatives of delivering an application on the Internet.

Back in the early days (about the time Al Gore was inventing the internet) if you wanted to have a web presence it meant that you had to acquire the hardware, bandwidth and technical support resources to run it from your own facility. While this model had many limitations, it wasn’t as difficult or costly for organizations which already had data centers and technical support personnel. However, this was a large barrier to entry for small, start-up organizations not only because of the hardware cost but also due to personnel costs (and backups). In addition, ensuring that backup power and alternate physical points of entry for multiple bandwidth providers was very difficult if not impossible to provide.

The next step, particularly during the internet boom of the late 1990’s, was the building of co-location facilities. Co-location facilities provided the services and facilities that were extremely expensive and difficult to build for an individual organization. This included security, cooling and humidity regulation, fire suppression, power regulation and emergency backup, adequate bandwidth provided by multiple providers entering the facility through physically dispersed entry ways and 24x7x365 on-site monitoring. Customers would bring their own hardware to the facility, install it in a cage or secure rack and then manage it either from on site or remotely. This was a huge step forward with regard to required capital costs and reliability.

This soon evolved into the facilities becoming more involved in managing hardware and applications, the advent of Managed Services, for which a premium fee was paid.

This was quickly followed by the renaissance of Open Source software and virtualization, including the availability of Virtual Private Servers (VPS), where one server could serve as what appeared to be a distinct server for several sites. Now a small organization could purchase some configuration of a VPS, or even a dedicated server, and very quickly and inexpensively have a fairly reliable and decent performing web presence.

It is this model, along with SaaS, that became one of the pillars of the Cloud Computing concept.

Amazon played a key role in the development of cloud computing by modernizing their data centers after the dot-com bubble burst, launching Amazon Web Services in 2006.

Now, a start-up has almost unlimited options from which to choose with regard to initial site hosting and configuration, starting at $4.95/month for a small VPS running up to $1,000’s/month for a firewalled, clustered, redundant configuration that is backed up daily, is fully managed with the almost instant ability to add bandwidth and processing power for peak periods, without having to absorb and justify the required capital expenditures.

The bottom line is that the options available to a start-up with regard to application delivery have greatly expanded, providing more options, more flexibility and at a lower cost, than ever before. However, it is important to continually watch those costs, when provided as there may come a time when other alternative delivery models may provide the same functionality at a lower cost.