Data Center Design Overview > Types of Server Farms and Data Centers (2023)

This chapter covers the following topics:

• Types of server farms and data centers

• Topologies of data centers

• Fully redundant Layer 2 and Layer 3 designs

• Fully redundant Layer 2 and Layer 3 designs with services

This chapter focuses on three key properties of data center architectures: scalability, flexibility, and high availability. Data centers are rapidly evolving to meet higher expectations for growth, consolidation and security. While traditional Layer 2 and Layer 3 designs have not changed dramatically in recent years, stringent uptime and service availability requirements, along with new technology and protocols, make design efforts more challenging and demanding.

The requirements for scalability, flexibility and high availability can be summarized as follows:

• ScalabilityThe data center must support rapid and seamless growth without major disruptions.

• FlexibilityThe data center must support new services without major infrastructure overhaul.

• High availabilityA data center should not have a single point of failure and should allow for predictable uptime (related to hard failures).

Scalability translates into the ability to support rapid growth in performance, the number of devices located in the data center, and the quantity and quality of services offered. Higher performance implies tolerance to very short-term changes in traffic patterns without packet loss and longer-term plans that show data center capacity growth trends.

Hosted appliance scalability refers to the ability to seamlessly add multiple ports to servers, routers, switches, and other service devices, such as server load balancers, firewalls, IDSs, and SSL offloaders. Higher density also includes slot density because the number of slots ultimately determines the potential growth of the system.

Flexibility translates into designs that can adapt to new service offerings without the need for complete architecture redesign or drastic changes beyond normal scheduled maintenance. The flexibility approach is a modular design where the features of the modules are known and the steps to add more modules are simple.

High availability translates into a fully redundant architecture where all possible severe failures are predictable and deterministic. This means that the failure of each potential component has a predetermined failover time and recovery time and that the worst possible failure scenario is still within the acceptable failover limits and within the requirements measured from an application availability perspective. This means that while the failure and recovery time of a network component should be predictable and known, the more important time relates to the user's perception of the time required to recover an application service.

Figure 4.1 provides an overview of a data center that as an object contains numerous building blocks and components of a larger corporate network architecture.

This book is mainly concerned with the engineering of application environments and their integration into the rest of the corporate network. Different types of server farms support application environments, but this book focuses on understanding, designing, implementing, and maintaining server farms that support intranet application environments. The actual engineering of different types of server farms Internet, extranet, and intranet server farms do not differ much from one type to another; however, their integration with the rest of the architecture is different. Design choices that differ for each type of server farm are a result of their main functional purpose. This leads to a specific location for their placement, security considerations, redundancy, scalability and performance. In addition to server farm concepts, a brief discussion of server farm types further clarifies these points.

Figure 4-1 Data center topology overview

As shown in Figure 4.1, there are three different types of server farms:

• s internet

• Extranet

• Intranet

All three types are found in the data center and often in the same data center facility, which is generally calledcorporate data centerfromcorporate data center. If the sole purpose of a data center is to support Internet applications and server farms, the data center becomes oneInternet data center.

Server farms are the heart of a data center. In fact, data centers are built to support at least one type of server farm. Although different types of server farms share many common architectural requirements, their goals differ. So the specific set of requirements for a data center depends on what type of server farm needs to be supported. Each type of server farm has its own set of infrastructure, security, and management requirements that must be met when designing a server farm. Although each server farm design and its specific topology may differ, the design guidelines apply to all server farms. Server farms are introduced in the following sections.

Internet server farms

As their name suggests, Internet server farms monitor the Internet. This means that users accessing server farms are primarily located somewhere on the Internet and use the Internet to reach the server farm. Internet server farms are then available to the wider Internet community and support business-to-consumer services. Internal users usually also have access to Internet server farms. Server farm services and their users depend on the use of web interfaces and web browsers, making them ubiquitous in Internet environments.

There are two different types of Internet server farms. A dedicated Internet server farm, shown in Figure 4-2, is built to support the large Internet-facing applications that support the core business. Typically, a company's core business is based on an online presence or online store.

In general, dedicated Internet server farms exist to support an enterprise's e-business goals. Architecturally, these server farms follow the data center architecture presented in Chapter 1, "Data Center Overview," but the details of each layer and the required layers are determined by the requirements of the application environment. Security and scalability are major concerns in this type of server farm. On the one hand, most of the users accessing the server farm are on the Internet, which poses a greater security risk; on the other hand, the number of likely users is very high, which can easily cause scalability problems.

A data center that supports this type of server farm is often called an Internet Data Center (IDC). IDCs are built by both companies to support their own e-business infrastructure and by service providers that sell hosting services, enabling companies to integrate e-business infrastructure into the service provider's network.

The next type of Internet server farm, shown in Figure 4-3, is built to support Internet-based applications with business Internet access. This means that the infrastructure that supports server farms is also used to support business users' Internet access. These server farms are usually located in the demilitarized zone (DMZ) because they are part of the corporate network, yet they can be accessed via the Internet. These server farms are called DMZ server farms, to distinguish them from dedicated Internet server farms.

Figure 4-2 Dedicated Internet server farms

These server farms support services such as e-commerce and are access portals for generic applications used by Internet and intranet users. Scalability considerations depend on how large the expected user base is. Security requirements are also very strict as security policies are aimed at protecting server farms from outside users while keeping the corporate network secure. Note that under this model, the corporate network supports the campus, private WAN, and intranet server farm.

Figure 4-3 Server farm DMZ

Intranet server farms

The evolution of the client/server model and the widespread use of web-based applications on the Internet created the basis for the construction of intranets. Intranet server farms are similar to Internet server farms in terms of ease of access, but are accessible only to internal company users. As described earlier in this chapter, intranet server farms contain most of the critical computing resources that support business processes and internal applications. This list of critical resources includes intermediate and mainframe systems that support a wide variety of applications. Figure 4-4 illustrates an intranet server farm.

Note that the intranet server farm module is connected to core switches that are part of the enterprise backbone and provide connectivity between the private WAN and the Internet Edge module. Users accessing the intranet server farm are located on campus and in the private WAN. Internet users generally do not have access to an intranet; however, internal users using the Internet as a means of transportation can access the intranet using Virtual Private Network (VPN) technology.

Figure 4-4 Intranet server farms

The Internet Edge module supports several features, including the following:

• Secure your business network

• Management of Internet access from the intranet

• Access control to Internet server farms

The data center provides additional security for additional data protection in the intranet server farm. This is achieved by applying security policies at the edge of the data center and at the applicable application layers when trying to strengthen communication between servers at different levels. The security design applied to each layer depends on the application architecture and the desired level of security.

The access requirements of business users dictate the size and architecture of server farms. The growing number of users, as well as the increased load imposed by rich applications, increases the demand for server farms. This requirement forces scalability to be a critical design criterion, along with high availability, security, and manageability.

Extranet server farms

From a functional perspective, extranet server farms sit between Internet and intranet server farms. Extranet server farms continue the trend of using web applications, but unlike Internet or intranet server farms, they are only available to a select group of non-Internet or intranet-based users. Extranet server farms are primarily available to business partners who are considered remote but trusted users. The main purpose of extranets is to improve communication between companies by enabling a faster exchange of information in a user-friendly and secure environment. This reduces time to market and business costs. Communication between an enterprise and its business partners, traditionally supported by dedicated connections, is rapidly migrating to VPN infrastructure due to ease of installation, lower costs and support for simultaneous voice, video and data traffic over the IP network.

As explained earlier, the extranet concept is analogous to IDC in that the server farm is located at the edge of the corporate network. Since the purpose of an extranet is to provide server farm services to confidential remote end users, there are special security considerations. These security considerations mean that business partners have access to a subset of the company's applications, but cannot access the rest of the company's network. Figure 4-5 shows an extranet server farm. Note that the extranet server farm is accessible to internal users, but access from the extranet to the intranet is prevented or highly secured. Usually, access from the extranet to the intranet is restricted by the use of firewalls.

When designing an extranet topology, many factors must be considered, including scalability, availability, and security. Dedicated firewalls and extranet routers are the result of a highly secure and scalable network infrastructure for connecting partners, but if you only have a small number of partners to deal with, you can leverage your existing Internet Edge infrastructure. Some partners require a direct connection or dedicated private connections, while others expect secure connections via VPN connections. Server farm architecture does not change whether you are designing Internet or intranet server farms. Design guidelines apply equally to all types of server farms, but design details are dictated by the requirements of the application environment.

Figure 4-5 Server farm extranet

The following section discusses the types of data centers that have been briefly mentioned in this section.

Internet data center

Internet data centers (IDCs) are traditionally built and managed by service providers, but companies with an e-commerce business model also build and manage IDCs. The architecture of enterprise IDCs is very similar to service provider IDCs, but scalability requirements are typically lower because the user base is typically smaller and has fewer services compared to SP IDCs that host multiple clients.

In fact, the architecture of the IDC is the same as that shown in Figure 4.2. An interesting consideration of enterprise IDCs is that, if the business model requires it, the facilities used by the data center may be located in the service provider's data center, but they remain under the control of the enterprise. This is typically done to reduce the costs associated with building a server farm and shorten time to market by avoiding building an in-house data center from the ground up.

The company's data center

Enterprise data centers support many different functions that enable different business models based on Internet services, intranet services, or both. As a result, support for Internet, intranet, and extranet server farms is not uncommon. This concept is illustrated in Figure 4-1, where a data center facility supports any type of server farm and is also connected to the rest of the corporate network—private WAN, campus, Internet Edge, etc. Supporting intranet server farms is still the primary purpose of corporate data centers.

Enterprise data centers are evolving, and this evolution is partly the result of new trends in application environments, such as n-tier, web services, and network computing, but mostly stems from the critical nature of data stored in data centers.

The following section discusses typical topologies used in data center architecture.