Server Is One Of The Main Hardware Components Of Computer Networks That Forms The Backbone Of Computer Networks. Typically, The Main Use Of Servers Is To Build Client-Server Networks.

Servers, like personal computers, are a set of hardware components such as main memory, storage space, CPU, power supply, and other components that facilitate users to access communication services. However, servers differ in appearance and processing power.

What is a server?

A server is hardware or software-driven device that is responsible for servicing applications and clients. The server operates in such a way that it receives and responds to client requests. Typically, in LANs, servers connect to switches or routers and provide organizations with important capabilities. For example, organizations are allowed to host websites or implement email servers.

An important point that we need to mention in this section is the concept of virtual servers. A virtual server is a part of a physical server that has the characteristics of a full-fledged server. More precisely, the operating system and virtual server applications are installed on the virtual server. To be more precise, a virtual server is a small part of a server that is the same as a building unit.

When you set up a dedicated virtual server, you can access all of the dedicated virtual server space. Another important term in this field is dedicated server. A dedicated server refers to a server specifically designed to host a company, and access to that server is only available to the company that provided it. In this case, the dedicated server will not share with any other client.

Server configuration hardware equipment

Servers are powerful processing systems used to provide various services. Theoretically, it is possible to use personal computers as servers. However, servers are not comparable to desktop computers due to the powerful processing power they provide. In addition, servers have some additional hardware than networked computers, which we will discuss below.

Server motherboard

Like a personal computer, servers have a motherboard, and other components are installed on them. However, the server motherboard is different from the PC motherboard. For example, it has a larger chassis, which allows more main memory slots to be placed on it, and thus the motherboard supports main memory. In addition, server motherboards can support two physical processors, are equipped with two network cards, and support up to 1 terabyte of main memory.

For example, if we look at the DL380 G10 series HP server, one of the 10th generation ProLiant servers, its motherboard supports two 12-core Intel Zeon Gold 5118 processors with a maximum frequency of 2.3 GHz. The motherboard of this server has 24 DIMM RAM slots and can support up to 1.5 terabytes of memory.

In addition, the DL380 G10 Prolet Server uses the S100i chip in the Smart Array section and can support up to 14 SATA hard drives. In addition, it has a p408-a Smart Array that sits directly on the server motherboard and features Support 3 SAS hard drives. The motherboard of this server is equipped with 4 network ports at a speed of 1 Gbps.

Server CPU

The main part of the server is the processor or CPUs on which they are installed. The CPU becomes important when the server is to be used for virtualization. If the server uses weak processors, the access speed to the virtual machines will reduce, or the virtualization software such as ESXi and Vcenter will not install on the server due to poor processing power.

When buying a server processor, the most important criteria should be the number of processing cores, Hyperthreading technology, amount of cache, the data transfer rate between the processor and hardware components, amount and type of main memory, clock speed, and clock speed maximum heat. Which produces Left (TDP), the type of socket that the processor supports, and so on.

CPU cooling

The main function of the heatsink is to dissipate the heat generated by the processor. The more workload applied to the processor, the more heat it generates. If a good heatsink is not installed on the processor, too much heat will disrupt the system’s performance and, in critical cases, cause the server to start, which is not an interesting phenomenon. Therefore, you should choose the right option when buying a heatsink processor.

Server main memory

Server main memory is significantly different from PCs, and PC memory should not use on servers. It is not bad to know that the description of the parameters that you have to observe to choose the main memory alone consists of several articles of several thousand words.

However, the most important thing to consider when purchasing the server’s main memory is the support for ECC technology, which is used to detect errors. This technology allows the memory to automatically detect and fix data transmission errors, a feature that personal computer RAMs do not have.

In addition, other technologies such as Advanced Error Detection and Advanced Fault Resiliency are other important criteria that must be considered when purchasing the server’s main memory. In general, the main memories of the server are classified as follows:

UDIMM:

Un Unbuffered memories that have low power consumption and also low capacity.

RDIMM:

Registered memory is a two-line system with stability between the DRAM module and the memory controller to increase memory reliability. This register buffers the control and address signals in the modules. These memories consume more power but ensure that the server is running smoothly. Therefore, they perform better than UDIMMs.

LRDIMM:

Load-Reduced two-line memories that are unstable and have buffer resistance. They are denser than RDIMMs and use a buffer chip to reduce and minimize the workload on the memory bus. These memories are a good choice for data centers because of their good performance.

Hard server

Hard disks are the main storage medium on servers, although solid-state drives (SSDs) will gradually replace hard disks. Equipped hard drives have a relatively shorter life than solid-state drives due to their mechanical process but are less expensive and offer more storage capacity. One thing to note about the server hard drive is the supported interface. The SAS and SATA interfaces on hard disks describe the speed and bandwidth that a hard disk provides.

Clothes drives provide faster access speeds than hard drives, and power consumption and latency are low. SDD drives are available in SFF 2.5 or 2.5-inch LFF. In addition, 2.5-inch NVMe PCIe and M.2 SSD solid-state drives are also available. Today, solid-state drives are architecturally divided into MLC, SLS, TLC, and QLC groups.

RAID controller

The main task of the RAID controller on hard disks is to establish a connection between the hard disk with the CPU and the operating system. The servers have the ability to support software and hardware read. The most important tasks of a RAID controller include monitoring input/output activity, monitoring key parameters, predicting system problems, term troubleshooting and automatic recovery, error management, and protecting hard drives from failure.

Reed control is the processor, cache, and drive for best performance in RAID. Another thing to note about the Reed Controller is that the Reed battery is present in servers such as the HP and can cause power outages and information to be stored on the flash in the event of a power outage.

network card

The network card is responsible for connecting the server to the network. Important components of a network card include the processor, main memory, EEPROM, and bus. The bandwidth provided by network cards is not the same. Of course, it is possible to increase the server’s network ports by installing a network card.

For example, the Ethernet 1GB 4Port 634025-001 Ethernet server network card supports the PCI Express 2.0 interface and is equipped with 4xRJ45-Ethernet 10base-t / 100base-tx / 1000base-t ports. This network card uses a Broadcom Bcm5719 processor and provides access to a transfer rate of 1 Gbps. In addition, it supports the SNMP remote management protocol. This network card is compatible with 8th generation servers.

Host Pass (HBA) Card

Host Host Adapter (HBA) is an integrated board that manages the input and output processing process and provides the physical interface between the server and the storage device under the network. Given that the HBA significantly reduces the server CPU workload associated with data storage and retrieval, it helps to improve server performance. Typically, these cards are installed on the PCI Express slot. These cards have speeds in the range of 8 or 16 Gbps and are equipped with one or two ports.

Power supply

According to the author, the power supply is the beating heart of any server. The power supply provides the required power to all components of the server. Some server manufacturers install two or more power supplies on the server to achieve more redundancy. Of course, each power supply alone can supply the power required by the server. If one of them fails, the next power supply automatically enters the circuit, preventing the server from crashing.

In addition, the server power supply architecture is such that it can replace without having to shut down the server. Power supplies have a shorter lifespan because they withstand a lot of work pressure. The most important thing to consider when choosing a power supply is the efficiency it provides. Typically, efficiency in the range of 92 to 96% is ideal.

Server cooling fan

Servers, like personal computers, are equipped with fans that are responsible for dissipating heat. The fans are responsible for dissipating heat inside the server. The better the heat dissipation performance of the fan inside the server chassis, the longer the life of the internal components.