The success or failure of a Web site (in particular, an e-commerce site) often depends on the site's ability to meet user demands for availability and quick processing of Web pages and resources. A well-known example of how the failure of hardware systems can affect a company is the experience that an auction company had with site crashes in 1999 and the consequences that those crashes entailed.
Click the Auction Site Case to read about what happened to this auction company.

Planning for scalability, failover strategies, and hardware failures, as well as ensuring that the network has effective security hardware, are critical to the success of a Web site. In addition, appropriate service agreements will help ensure a successful operation.

Scalability is the ability of the system to grow and be able to handle more clients and transaction volume as needed. Planning for scalability ensures that your investment in hardware will not just go down the drain as your business grows because the site will be able to handle current and future user needs. For that reason, when you plan for scalability you must focus on:

1. Current volume (or expected volume for the immediate future)
2. Projected volume over time
3. Scalability solutions involve various components of the network, including servers and other hardware.

Failover is a strategy for dealing with server failure by having other servers take on the load. In the Requirements Definition, the Web team has defined the up-time requirements. Most sites (again, in particular, e-commerce sites) require high fault resilience to achieve those up-time requirements. In other words, failover strategies enable the site to provide the highest levels of uninterrupted service.
There are a variety of failover approaches that a Web team can take. The final choice depends on up-time requirements and the available budget.

It is important to plan for the possible failure of key site hardware. The design should include supplemental equipment to mediate hardware failure. This equipment may include: Hot swappable disks: disks that can be safely removed and replaced from a server without turning the server off.

RAIDs (Redundant Arrays of Inexpensive Disks): a set of techniques for combining disk drives into an array of drivers. Data is written across a number of drives, thus reducing the loss of data should one disk fail. There are several levels of RAID that can be implemented to provide data protection in the event of hardware failure (such as striping data across drives, or mirroring the data onto two drives).

Web sites may also fail due to outside attacks by hackers. A firewall ^[1]provides the primary line of defense against attackers.

Selecting a firewall requires factoring in a number of considerations, including cost, complexity, performance needs, the level of security required, as well as the company's existing security policy requirements for monitoring access control and authentication.

An important strategy in managing hardware risks that is often overlooked is to pay close attention to service agreements. Service agreements vary in terms of the guarantees of up-time and turnaround. The more critical the need for 100% up-time, the more expensive the service agreement is likely to be. In some cases, a manufacturer's representative(s) may be on site or available 24 hours, 7 days a week. In others, the system may simply alert the administrator via a pager of a system problem. The administrator then contacts the appropriate personnel as soon as possible. You must weigh your needs and match them with your budget, but you should not neglect this aspect of your strategy since a lot will depend on these agreements when you're in trouble.
In the next lesson, you will review what you have learned in this module about planning for a successful network architecture.

Click the Quiz link below to check your knowledge of connectivity and security hardware and the problems associated with it.
Connectivity Security Hardware - Quiz

[1] Firewall: A system that controls access between two or more networks. With a firewall, companies can control employee Internet access and access to company computers from outside the corporation. A firewall is an extra layer of security built into computer systems set up at the Web site or at the ISP's site that can be used to monitor and filter both incoming and outgoing requests.