Content Delivery Networks (CDN) is an inevitable part of any online business. Since this technology becomes so widespread and popular, more people want to understand what it is exactly. Read this article to learn the basic information about CDN and even more.

What is a Content Delivery Network?

A Content Delivery Network (CDN) is a network of servers distributed around the world that ensure faster delivery and better availability of content. Data is replicated throughout the CDN, and exists in several places at once. The client gets the copy of the data located near to him. Thus, all clients access different servers, which prevents bottlenecks.

In other words, CDN improves user experience and makes network utilization more efficient. Many media companies and e-commerce vendors benefit from CDN to transfer content to their clients. In its turn, CDN pays carriers and operators for hosting servers in their data-centers.

How does CDN work? First, it keeps content distributed on several data-centers so that it was closer to end users and faster to transfer. Server optimization should be done on the basis of content delivered to the user. Together with improving performance, CDN offloads traffic delivered from provider's infrastructure, what saves costs for content provider. With CDN, you also ensure protection against DDoS attacks, absorbing the malicious traffic with the help of sponge PoPs.

Why distance matters?

Location of servers and end-users is a very important aspect. Traditionally, a web-site is hosted on one server, which means that users from all corners of the world are connected to it. If, for example, a server is places in the USA, and a web-site is accessed by a person from China, latency cannot be avoided - data will have to be transferred via miles and miles of cables. You cannot offend against the laws of physics: the more content has to travel, the worse users' experience may be.

CDN architectures

Most CDN architectures include the following components:

- Delivery Nodes is what delivers data to customers. It has caches that run on one or several applications. The one that is the closest to the end user is deployed.
- Storage nodes provide data to caches, and can be used in a hierarchical model to make up for tiered caching and protect origin servers. They are useful in cases when pre-published content is required instead of content demanded from the origin server.
- Origin node is a source of content that can be used within operator's network, e.g. within user's infrastructure. It is recommended to have a number of origins for scale and resilience.
- Control node is designed to provide management, routing and control of CDN components.

As a rule, CDN nodes are located in different places, often - in multiple backbones. Why is that beneficial? First, bandwidth expenses are reduced, secondly, page load time shortens. That means that the availability of content globally is better. The number of nodes and servers comprising a CDN is different - it depends on the architecture. Sometimes there can be thousands of nodes and servers deployed on many PoPs. In other cases, a small number of PoPs is enough.

Types of content delivered

There are three main types of content, and CDN service can boost the transfer of all them:

- Dynamic content is the one that is generated in real time by a web server with the help of several programming languages (php, ruby, java, etc.);
- Static content is usually not changes so often - it is generated once and altered only when needed. These are images, JavaScript, text, CSS and other files.
- Streaming content is a video or audio file that is played via web browser control.

Each type of content varies in size. The time it takes for each piece of data to travel from the server to the user is called "latency", and is measured in milliseconds. The faster web content is, the less is latency. In mobile browsing, this is a more serious problem, which is why mobile content delivery requires more effective optimization strategies. Each additional second of loading may cause 7-10% drop of conversion rates. CDN has been developed to solve all these issues. It reduces latency and helps to deliver content easier and faster.

History of Content Delivery Networks

The Internet is permanently developing and changing, and new types of data and content are introduced eventually. As soon as this technology has become available for most people, the problem of sending huge amounts of data has appeared. CDN was invented to help with optimization of the speed of web.

The basis of CDN technology was created about 20 years ago, and it is still used to drive the force of delivery. After that, there have been significant advances reached by commercial and scientific developers. CDN can be considered as one of the technologies that seriously impact web experience.

Since CDN was introduced, the market experiences an enormous boost of broadband delivery and video and audio streaming in the Net. The life cycle of CDN is the following:

1. The first (pre-formation) period was market by introduction of server farms, hierarchical caching, proxy deployment and improving of web-servers. That was the benchmark of infrastructure development and caching, which served as basis for CDN.
2. After than the first generation of CDNs appeared, and it was mostly about delivery of static and dynamic content (only these content types existed then). At those times, principle mechanisms are replicas, intelligent routing and methods of edge computing. Information and applications were split across servers.
3. The second generation of CDNs was focused on VoD (video on demand) - streaming audio and video on different sites. It gave way for mobile content delivery, which was a real breakthrough then. This generation of content delivery networks for streaming also used P2P and cloud techniques. However, the technology was still on the stage of its origin.
4. The third generation was designed so that the systems could be driven by regular users and individuals. Self-configuration appeared to be a very helpful new mechanism together with autonomic content delivery. This stage was about improvement of users' experience.

Having been evolving for about a decade, CDN always adapts to the needs of customers and changing technologies. It's not surprising: technologies are constantly developing, and we have to adjust. It is difficult to predict which way the CDN technology will alter and evolve. Until recently, development was connected with bandwidth pressure and video streaming. But with cloud computing, CDN involves all layers: software, infrastructures, platforms, business processes as a service, etc.

How Performance Optimization is Made?

When it comes to web-site hosting, it should first and foremost be aimed at perfect user experience. This is why many web projects feel enormous pressure: from one side, users have great expectations concerning interactivity and media, from other side - Internet is growing, and it gets more and more difficult to create a user-friendly web-site.

Optimization of website and its performance implies improvement of speed of its web elements (images, texts, other files) which are displayed or downloaded by customers' web browsers. Performance directly influences web-site conversion rates, brand perception, clients' loyalty and their engagement. Since about quarter of e-commerce sales are done via phones, mobile connectivity cannot be ignored. In many cases, users who are made to wait more than 4 seconds for a web-site to load will simply abandon it.

So how CDN is connected with performance issues? To put it simply, it improves it by speeding up web-sites. Here are the main reasons why site developers should strive to boost performance and speed:

1. Users' satisfaction. Long loading time increases bouncing rates (when a user visits your web-site and leaves it instead of surfing). This aspect is crucial for ranking, and is always taken into account by searching engines. Bounce rate is considered as telltale sign of web-site effectiveness. Taking into account the level of technology development today, a simple web-site should load within a couple of seconds. Loading time is what initially defines satisfaction or discontent of your users.
2. Improved search rankings. Since 2010, Google started considering web-site loading time when ranking web-sites. This is a prioritized factor among many, so web-site speed should always be taken into account. High loading speed helps you to stay in the top of search results - this is an essential condition for being spotted by your potential customers. No matter what exactly your clients like - attractive interface, transparent privacy policy, great bargains, or useful information - they would want to get it quickly. If you reduce web-site speed even for a second, it will already convert much more visits into real orders.
3. Attraction of mobile users. They tend to be more patient because of always dealing with slow Internet. But if you consider the possibilities of modern technologies, loading time for mobile connection can and should be reduced. As you can guess, things are a bit different for tablets and phones, but there is a lot that can be done to solve some problems of web-site loading time. Optimization of mobile pages is more important than ever, because the number of mobile uses steadily grows together with the number of devices. Averagely, 250 ms of page load time will be enough to keep your customers from going to a competition. If you prefer CDN services over usual hosting, site content will be delivered from the server closest to the end user. That excludes the possibility to invest in infrastructure and software, being a low cost solution.

CDN is a necessary condition of fast connection and great user performance. This technology is at the peak of its development, and will be deployed by developers globally further on.