Blade server and rack server advantages are completely compared

We often see the server, from the appearance type can be divided into three types, namely tower server, rack server and blade server. Blade servers and rack servers are becoming more and more popular among users due to factors such as limited space in enterprise rooms. What are the characteristics of blades and rack servers? Which article is better for you?

1. Rack Server and Its Features

Rack Server refers to a server that can be directly installed into a standard 19-inch cabinet. Usually such a server looks like a switch in size. Therefore, the rack server is actually a product under industrial standardization, and its appearance is designed according to a unified standard, and is used uniformly with the cabinet to meet the intensive deployment requirements of the enterprise. The main function of the rack server is to save space. Since multiple servers can be installed in one cabinet, it can occupy less space and is easy to manage. The height of a typical cabinet is 42U (1U = 1.75 inches or 4.4 cm), the width of the rack server is 19 inches, and most rack servers are 1U-4U high.

The advantage of the rack server is that it takes up a small space and is easy to manage. However, due to internal space limitations, the scalability is limited. For example, a 1U server has only one or two PCI expansion slots. In addition, the heat dissipation performance is also a problem that needs attention, in addition to the need for cabinets and other equipment, so this kind of server is mostly used by large enterprises with a large number of servers, and many enterprises use this type of server, but deliver the server to A dedicated server hosting agency to host, especially for servers on many websites today.

Second, blade server and its features

The blade server is a high-availability, high-density, low-cost server platform designed for special application industries and high-density computer environments. The structure is a large main body chassis, and many "blades" can be inserted inside. Each of the blades is actually a system motherboard, similar to a separate server, which can boot its own operating system through the local hard disk. Each blade can run its own system, serving a different set of users, without correlation. Moreover, these motherboards can also be combined into a server cluster using system software. In cluster mode, all blades can be connected to provide a high-speed network environment, share resources, and serve the same user base. Inserting a new blade into the cluster improves overall performance. And because each blade is hot-swappable, the system can be easily replaced and minimized maintenance time.

Based on the server functions that need to be undertaken, blade servers are divided into server blades, network blades, storage blades, management blades, Fibre Channel SAN blades, expansion I /O blades and other different functions of the blade server. There are two recognized features of the blade server. One is to overcome the shortcomings of the chip server cluster and become the terminator of the cluster; the other is to achieve cabinet optimization.

Three, the advantages of blade server and rack server full comparison

The current blade server products will be driven by multi-core and low-power technology, from the initial pursuit of high-density blade servers. A new generation of blade products that emphasize overall performance and productivity. Blade servers will focus on higher density, agile deployment and maintenance, comprehensive monitoring management, high scalability, high availability, and become a mature mainstream product that goes hand-in-hand with rack-mounted servers. At present, blade servers have opened up breakthroughs in many industries, market demand is constantly being stimulated, and the future of the blade server market attracts attention.

At present, as a new star in the server field, the market prospects brought by such high-density blade servers have been favored by major server vendors, including of course international brands such as IBM, HP, SUN, and DELL. And domestic famous brands such as Lenovo, Inspur, and Shuguang. Let's compare the advantages of blade servers and rack servers.

1. Blade server and rack server from space density

The blade server is more space-saving than the rack server. In fact, the blade server is the space density occupied by the rack server. Once again increased by 50%. On the premise that the cabinet system is configured, the 1U rack optimization server system is transplanted to the blade server, occupying only 1/3 to 1/2 of the original space. In a standard rack-mounted environment, the processing density of the blade server is increased by four to five times. For example, in a high-density computing server environment that handles 1024 nodes, if all servers with 1U configuration are used, 24 cabinets are needed, which does not include the cabinet space occupied by the Ethernet switching hub, and 8 "blades" are inserted. The blade server requires only 9 cabinets, but it includes the cabinet space of the Ethernet switching hub.

2. From the perspective of cabling and management of the equipment room, the blade server and the rack server are

The blade server is wired in the equipment room as long as the network cable and power cable are uniformly distributed, and no artificial wiring is required between the blade servers. The rack server should separately wire the network cable and power cable of each server. If multiple 1U servers are installed in a 42U cabinet, the wiring behind the cabinet is very much, which seems messy. If we need 10 highly reliable 1U rack servers and need KVM, network and management functions, then we need 20 power supplies (redundant power supplies), 10 KVM cables, and at least 20 network cables. (Each server requires one management line and one network cable), which does not include cables for connecting external storage and other devices. The cable routing of the blade server solution is significantly better than the 1U rack solution.

From a TCO perspective, blade servers are easier to manage, provide more processing power in a smaller space, and cost less. Because many components can be shared, unlike 1U servers. For example, in the above example, after using the blade server, we no longer need 10 DVD drives, 10 different remote management modules and 20 power supplies.

3. Flexibility in terms of blade servers and rack servers

Most blade servers are less flexible than rack servers. For example, in high-performance database applications, blades cannot be equipped with disk arrays through external RAID cards. In addition, if the user needs a large-capacity database server, but does not want to build through the cluster, the blade will not be able to do so. Because rack servers with 16 memory slots are readily available on the market, blade servers can only support 4-8 memory slots. It can be seen that the flexibility of the blade server is relatively weak, and even if there is a suitable product, the price is extremely high.

In most cases, the blade supports a 2.5-inch hard drive, which is not only more expensive than a 3.5-inch hard drive, but also has a lower performance, which is also thanks to the "blade pursuit of high density." Today, a typical blade can only be installed with at least 140GB hard drives, while many rack products can be equipped with 2TB (4X500GB) storage capacity.

4. From the perspective of procurement costs, blade servers and rack servers

theoretically, due to the reduction of many redundant unnecessary components such as DVD, power supply, and KVM and network lines Cable, blade server purchase costs will be lower than the same number of rack servers. Blade server vendors each have their own proprietary blade architecture, and because of their monopoly, the price of IBM, HP and DELL blade products remains high.

People think that the cost of a fully equipped blade server is much lower than that of the most expensive rack server from the same vendor. But in fact, if the user takes the approach of "increasing the blade in the chassis", it will be found that the rack server is more cost-effective. Because most blades and related products are more expensive than the corresponding rack products, the average blade chassis costs $4,000-$8,000, and the blades are also more expensive than the 1U racks, at least not cheap.

5. From the perspective of scalability and standards, blade servers and rack servers

The advantages of 1U rack servers are that they use standard server design technology, standard peripherals, and standard interfaces. With RAID function and redundancy function, it can run independently and undertake tasks. In terms of scalability, the rack server has poor expansion performance in the chassis because of the small space of the chassis, but it can be very effectively expanded by means of a server cluster or an external expansion cabinet. Blade servers are innovative in scaling up and scaling out. Adding a new server typically involves simply inserting a new single-processor or multi-processor blade into the open bay of the chassis. The blade server can be plugged into an expanded infrastructure. In addition, the option module inside the chassis allows us to add the ability to share once externally connected. The modular design of blade technology enables rapid expansion.

The blade products of IBM, HP and DELL are still not compatible, there is no standardization, and the different blade chassis standards of the same manufacturer are different. The rack server is a unified standard. IBM's 2U server can be installed in DELL's 42U server cabinet.

6. From the perspective of reliability and maintenance, the blade server and the rack server are

when the common 1U rack server solution is adopted, the connection points generated by a large number of power lines and network cables are formed. After a large number of potential "problem points" were replaced by blade server chassis and blade server solutions, the original network and power connection points were reduced, and the system reliability was increased.

All key components of the blade server are redundant or hot-swappable, including cooling systems, power supplies, Ethernet controllers and switches, midplane and backplanes, hard drives, and service processors. Removing the server for repair simply means pulling the blade server out of the chassis, which is as simple as removing a hot-swappable hard drive. Advanced blade server systems provide an intelligent way to achieve highly sensitive repairs, and advanced diagnostics guide maintenance personnel to locate faulty components directly for fast and efficient recovery, with some blade servers not even having a single point of failure. Rack server maintenance is more complicated than blades.

In short, rack servers and blade servers are functional servers that focus on a specific application. The rack server can meet the needs of user applications. The integration of the blade server is too dense, and the standard and heat dissipation have not been completely solved. The blade server is suitable for a large number of machines to form a cluster for users to use. Therefore, the blade is used. Up to now, the server has not taken an absolute advantage compared to the rack.