A computer network is called. Setting up Windows network tools. “What is a computer network? Types of networks"

Definition of a computer network. Types of computer networks.

Networks – a type of distributed IS. A system that collects, stores and processes information. Types of distributed systems: Network -> cluster -> multiprocessor system.

A cluster is a system in which nodes are physically isolated from each other, but there are special means to connect them. Buses are used to transmit data.

Nodes are specific computers.

Network composition:

1) Computers (hosts) are consumers and generators of information;

2) Network equipment: a) hubs, b) bridges, c) switches,

d) repeaters, e) routers, f) firewalls, g) network cards,

h) network cables, i) modems (ADSL, Wifi, Cable, dial-up are outdated due to low speed).

Computer network (Computer Network) - a set of computers connected by communication lines and running special software.

Server– a computer or program that provides a specific service.

Client– an application program, a consumer of services or information transmitted by the server. Home purpose of the association computers on the network - providing users with access to various information resources distributed across these computers and their sharing.

Important characteristic any computer network - latitude of territory which it covers. The breadth of coverage is determined by the mutual distance of the computers that make up the network and, therefore, affects the technological solutions chosen when building the network.

Based on the breadth of coverage, there are classically 2 types of networks:

- Local LAM networks, the computers of which are concentrated in relatively small areas (usually within a radius of 1-2 km). Quite expensive and high-quality technologies are used to build them, which ensures high speed exchange of information between computers.

- Global WAM are networks designed to connect individual computers and local networks located at a considerable distance (hundreds and thousands of kilometers) from each other. In global networks, lines that already exist and were not originally intended for building computer networks (for example, telephone or telegraph lines) are often used. In this regard, the data transfer speed in such networks is significantly lower than in local ones.

They also highlight:

- City networks (MAN). Such networks are designed to ensure the interaction of computers and/or local networks dispersed throughout a large city (usually within a radius of 100 km), as well as to connect local networks to global ones. To build such networks, sufficiently high-quality digital communication lines are used, allowing interaction at relatively high speeds compared to global networks.

- Internet – a network that brings together local, city and global networks of the entire planet.

By type of organization:

- The local network is a network in which computers are located compactly, either in one or in several adjacent rooms.

- Peer-to-peer network is a network in which there are no dedicated servers, all computers have equal rights and are both clients and servers at the same time.

- Heterogeneous network is a network to which computers with different operating systems are connected.

Definition of a computer network. Types of computer networks. - concept and types. Classification and features of the category "Definition of a computer network. Types of computer networks." 2017, 2018.

What is a network

A computer network is a collection of PCs and other devices connected together through network cables so that they can communicate with each other to share information and resources. Networks vary in size: some are located within a single office, others span several buildings and even the entire globe.

When creating networks, they are most often used Ethernet technologies And Fast Ethernet. Several technologies can be used in one network. Ethernet and Fast Ethernet networks function similarly; the main difference is the data transfer speed.

How the network works

Information is transmitted in packets. Each packet contains the address of the sending and receiving devices, which allows it to reach its destination.

Ethernet and Fast Ethernet networks use the CSMA/CD (Carrier Sense Multiple Access with Collision Detection) protocol. This protocol allows only one device to transmit data at a time. If two devices try to transmit information at the same time, a collision occurs, which is detected by the transmitting devices. Both devices stop transmitting and wait until they can start transmitting data again. This mechanism is similar to a conversation between two people: if both start talking at the same time, they stop and then one starts talking again.

Networking benefits

On the network, both information and resources can be shared or, as they say, shared by users. This has a number of advantages:

• you can use the same peripheral devices, such as printers, modems, scanners, etc. (for example, there can only be one printer on the network);
• You can transfer data without using floppy disks. Transferring files over the network saves time spent writing and reading data from a floppy disk. In addition, there are no restrictions on the size of copied files;
• you can centrally use important computer programs, for example accounting. Often users need to be able to access and work with the same program at the same time. For example, the ticket sales program should be uniform to prevent reselling;
• it is possible to provide automatic backup important files. The backup program starts automatically, saving time and ensuring the safety of your files.

Network components

A small network usually consists of:

• PC and peripheral devices such as printers;
• network adapters for PCs and network cables;
• network equipment such as hubs and switches that connect PCs and printers;
• network operating system, such as Windows NT or NetWare.

In addition, other equipment may be required.

In order for your PC to be used on the network, you must install network adapters. Some PCs come with a pre-installed network adapter. The network adapter must be compatible in speed with the hub to which the PC is connected. Yes, network Ethernet adapter corresponds to an Ethernet hub, and a Fast Ethernet network adapter corresponds to a Fast Ethernet hub.

Hubs

The term "hub" is sometimes used to refer to any network device that serves to connect PCs to a network, but a hub is actually a multiport repeater. These types of devices simply transmit (repeat) all the information they receive - that is, all devices connected to the hub ports receive the same information.

Hubs are used to expand the network. However, over-reliance on hubs can result in a lot of unnecessary traffic being sent to network devices. After all, hubs transmit traffic to the network without identifying the actual destination of the data. PCs that receive data packets use the destination addresses present in each packet to determine whether the packet is intended for them or not. On small networks this is not a problem, but even medium-sized networks with heavy traffic should use switches that minimize the amount of unnecessary traffic.

Switches

Switches monitor network traffic and manage its movement by analyzing the destination addresses of each packet. The switch knows which devices are connected to its ports and routes packets only to the necessary ports. This makes it possible to simultaneously work with several ports, thereby expanding the bandwidth.

In this way, switching reduces the amount of unnecessary traffic that occurs when the same information is sent to all ports.

Switches and hubs are often used on the same network; hubs expand the network by increasing the number of ports, and switches break the network into smaller, less congested segments.

When to Use a Hub or Switch

In a small network (up to 20 workstations), a hub or group of hubs can handle network traffic quite well. In this case, the hub simply serves to connect all users on the network.

On a larger network (about 50 users), it may be necessary to use switches to divide the network into segments to reduce the amount of unnecessary traffic. If you use a hub or switch with indicators showing the degree of network congestion, then by analyzing their readings, you can draw certain conclusions. So, if the traffic is constantly high, you should use a switch to divide the network into segments. If you add new hubs to your network, you must follow rules that limit the number of hubs that can be directly connected to each other. The use of switches allows you to increase the number of hubs used in the network and thereby expand the network.

Network formation rules

Ethernet and Fast Ethernet Rules

When forming a network of several devices, you must follow a number of rules related to:

• the number of hubs that can be connected to each other;
• the length of the cable used;
• the type of cable used.

These rules are similar for Ethernet and Fast Ethernet. If you are dealing with hubs that support two types of connections - Ethernet and Fast Ethernet, then you must use Ethernet or Fast Ethernet rules depending on the type of equipment connected to the hub. If you connect two hubs together, then there should be a Fast Ethernet connection.

When you need to connect more users to your network, you can simply use another hub by connecting it to your existing network equipment. Hubs work differently than other network equipment. They simply transmit the information coming to them to all other ports. There is a limit to the number of hubs that can be connected together because a large number of hubs causes the network to be susceptible to collisions.

In 10Base-T Ethernet networks, the maximum number of hubs located in a row should not exceed four.

The problem can be solved by placing one switch between the hubs. As you know, switches divide the network into segments. In this case, the switch should be positioned so that there are no more than two hubs between the PC and the switch. It is this structure that meets the requirements of Ethernet and guarantees correct operation of the network.

Rules for Fast Ethernet on Twisted Pair

The maximum number of hubs in one branch is two.

100Base-TX requires Category 5 twisted pair cable. The maximum cable segment length is 100 m. The total length of twisted pair cable running through directly connected hubs must not exceed 205 m.

Connectors and cables

Why is twisted pair replacing coaxial cable?

Coaxial cable

Twisted pair and coaxial cable are Various types cables that are used to connect computer network equipment.

Coaxial cable began to be used in networks before twisted pair. Coaxial cable networks are created by combining T-sections into one long segment. The two free ends of the segment are terminated by terminators. PCs are connected to one end of the T-section. Data is transmitted along the entire segment and reaches all devices included in the segment. For the network to function, the entire segment must remain intact. This means that if any section of the cable is damaged or disconnected, the network will not work. During the process of upgrading the network (for example, when adding new PCs), a segment is broken, which makes the network temporarily inoperable.

Coaxial cable can only be used for Ethernet networks.

Twisted pair cable

Twisted pair cable is easier to use and much more flexible than coaxial cable. Therefore, in most networks, twisted pair cable is used as the physical transmission medium. Small twisted pair networks usually use a central device - a hub or switch - to which all PCs are connected via twisted pair. This device distributes information between PCs that are connected to it.

The twisted pair cable is very flexible and has easy-to-use connectors that fit easily into the ports network equipment, PCs and printers. If a twisted pair cable is damaged, only the device it connects to the network will be blocked. All other devices remain operational. Upgrading the network (for example, adding new PCs) is very easy, and this process does not affect the operation of other devices. Category 5 cable can be used for Fast Ethernet networks. In addition, the use of Category 5 cable will allow you to move from Ethernet networks to Fast Ethernet networks.

How can you establish a network connection using a coaxial cable?

Previously, coaxial cable was most often used as a transmission medium in Ethernet networks. In order to ensure the transition to twisted pair, you should use hubs with two types of ports - twisted pair and coaxial cable.

If you have one of these hubs, establishing a network connection using coaxial cable should be fairly easy. All you need to do is connect existing network with the corresponding hub port. In addition to connecting to a coaxial cable, these hubs also allow connections via twisted pair.

Expansion and modernization

Let's consider network expansion issues related to an increase in the number of users.

In what cases is it necessary to expand or upgrade the network? There are at least three reasons:

• more ports required;
• wider bandwidth required;
• peer-to-peer networking has become too complex.

More ports required.

In cases where the number of network users increases, you can simply add another hub by connecting it to the existing one. As a result, the required number of additional ports will appear.

To pair devices on Ethernet networks, an MDI interface is used, which regulates the connection rules. Most hub and switch ports are MDI-X ports, which use a standard twisted pair cable to connect to a PC. Some ports may belong to the MDI/MDI-X category. This means that their operating mode is selected using a switch. In order for a connection between two devices to work, the receiving lines of one device must be connected to the transmitting lines of the other. To connect two MDI ports or two MDI-X ports, you must use a so-called crossover cable - a cable with the transmitting and receiving lines crossed. Let's formulate the most commonly used rules for pairing devices:

1. To connect the MDI/MDI-X port (set to MDI mode using a switch) of one device to the MDI-X port of another, you need a standard twisted pair cable.
2. To connect two MDI-X ports, you need a twisted pair crossover cable.

More bandwidth required.
Hub-based Ethernet networks are the ideal solution for most small networks. However, if the Ethernet network constantly operates under heavy loads, then you can:

• add an Ethernet switch. If there are more than 25 users on the network or most users have regular Ethernet adapters in their PCs, then adding an Ethernet switch will allow you to divide the network into less loaded segments;
• go to Fast Ethernet. If a lot is transmitted over the network large files(for example, graphics), then moving to Fast Ethernet will provide 10 times more bandwidth. This will speed up file transfers and other network operations.

Note that the transition to Fast Ethernet will require network adapters of the Fast Ethernet standard. If you are not planning to upgrade your entire network at once, we recommend using auto-sensing hubs. These devices provide auto-configuration of Ethernet/Fast Ethernet ports, allowing you to connect old Ethernet equipment and new Fast Ethernet equipment to the hub.

Peer-to-peer networking has become too complex.

In a number of cases, difficulties arise associated with the growth of a peer-to-peer network:

• If there are many shared folders or files, then it is inconvenient to control their location and access rights.
• If shared folders and local printers are accessed frequently, this will slow down the performance of the PCs to which they are connected.

Client-server technology

Very often it is advisable to move from a peer-to-peer network to a network based on client-server technology, the use of which allows for more efficient use of LAN capacity. In this case, the application is divided into two parts: client and server. One or more of the most powerful computers on the network are configured as application servers; they run application servers. The client parts run on workstations; It is on workstations that requests to application servers are generated and the results obtained are processed.

There are networks with one or more dedicated servers. In such networks, it is the server resources, most often disk memory, that are available to all users. Servers whose shared resource is disk memory are called file servers. We can say that the server serves all workstations. A file server is usually used only by a network administrator and is not intended for solving application problems. Therefore, it can be equipped with an inexpensive, even monochrome display. However, file servers almost always contain multiple high-speed drives or even a RAID array. The server must be highly reliable, since its failure leads to the shutdown of the entire network. The file server usually has a network operating system installed: most often it is Windows NT, NetWare or Linux.

Workstations are installed with a conventional operating system, such as DOS, Windows or Windows NT. The workstation is individual workplace user. Unlike a peer-to-peer network, the user is the full owner of all resources on a workstation. At the same time, the file server resources are shared by all users. A computer of almost any configuration can be used as a workstation. But ultimately it all depends on the applications that a given computer uses.

ComputerPress 10"1999

Computer network (Computer Network) is a set of computers connected by communication lines and running special software.

Under communication line usually understand the totality technical devices, and the physical environment that ensures the transmission of signals from the transmitter to the receiver. IN real life examples of communication lines include sections of cable and amplifiers that provide signal transmission between switches telephone network. Communication channels are built on the basis of communication lines.

Communication channel usually refers to a system of technical devices and communication lines that ensures the transfer of information between subscribers. The relationship between the concepts of “channel” and “line” is described as follows: a communication channel can include several heterogeneous communication lines, and one communication line can be used by several channels

Communication lines and channels

The main purpose of connecting computers into a network is to provide users with the ability to access and share various information resources (for example, documents, programs, databases, etc.) distributed across these computers.

An important characteristic of any computer network is the breadth of the territory it covers. The breadth of coverage is determined by the mutual distance of the computers that make up the network and, therefore, affects the technological solutions chosen when building the network. Classically, there are two types of networks: local networks and wide area networks.

Local networks

TO local networks(Local Area Network, LAN) usually refers to networks whose computers are concentrated in relatively small areas (usually within a radius of 1-2 km). A classic example of local networks is the network of one enterprise located in one or more nearby buildings. The small size of local networks allows the use of fairly expensive and high-quality technologies for their construction, which ensures high speed information exchange between computers.

The local network

Global networks

Global networks(Wide Area Network, WAN) are networks designed to connect individual computers and local networks located at a considerable distance (hundreds and thousands of kilometers) from each other. Since the organization of specialized high-quality long-distance communication channels is quite expensive, global networks often use existing lines that were not originally intended for building computer networks (for example, telephone or telegraph lines). In this regard, the data transfer speed in such networks is significantly lower than in local ones.

Global network

City networks

Not so long ago, another one was added to the two indicated types of networks - the so-called city ​​networks(Metropolitan Area Network, MAN). Such networks are designed to ensure the interaction of computers and/or local networks dispersed throughout a large city (usually within a radius of 100 km), as well as to connect local networks to global ones. To build such networks, sufficiently high-quality digital communication lines are used, allowing interaction at relatively high speeds compared to global networks.

City network

Internet

Regardless of what territory the network covers, what technological solutions underlie its organization, there are general principles of network interaction to which the functioning of the network must be subject. It is the development of such general principles contributed at one time to the emergence of the Internet as a unified network (sometimes the term “hypernetwork” is even used), which brought together local, city and global networks of the entire planet.

Internet as a connected network

The following sections highlight modern concepts of network interaction, as well as their practical implementations that technically ensure the functioning of the Internet.

• Tutorial

Hi all. The other day I had an idea to write articles about the basics of computer networks, analyze the work of the most important protocols and how networks are built in simple language. I invite those interested under cat.

A little off-topic: About a month ago I passed the CCNA exam (with 980/1000 points) and there is a lot of material left over the year of my preparation and training. I first studied at the Cisco Academy for about 7 months, and for the remaining time I took notes on all the topics that I had studied. I also advised many guys in the field of network technologies and noticed that many stumble on the same rake, in the form of gaps on some key topics. The other day a couple of guys asked me to explain what networks are and how to work with them. In this regard, I decided to describe the most key and important things in as much detail and in simple language as possible. The articles will be useful to beginners who have just embarked on the path of learning. But perhaps experienced system administrators will also highlight something useful from this. Since I will be taking the CCNA program, this will be very useful for those people who are preparing to take the test. You can keep articles in the form of cheat sheets and review them periodically. During my studies, I took notes on books and read them periodically to refresh my knowledge.

In general, I want to give advice to all beginners. My first serious book was Olifer’s book “Computer Networks”. And it was very difficult for me to read it. I won't say that everything was difficult. But the moments where it was explained in detail how MPLS or carrier-class Ethernet works were stupefying. I read one chapter for several hours and still a lot remained a mystery. If you understand that some terms just don’t want to pop into your head, skip them and read on, but under no circumstances discard the book completely. This is not a novel or an epic where it is important to read chapter by chapter to understand the plot. Time will pass and what was previously incomprehensible will eventually become clear. This is where your “book skill” is upgraded. Each subsequent book is easier to read than the previous book. For example, after reading Olifer’s “Computer Networks,” reading Tanenbaum’s “Computer Networks” is several times easier and vice versa. Because there are fewer new concepts. So my advice is: don't be afraid to read books. Your efforts will bear fruit in the future. I’ll finish my rant and start writing the article.

Here are the topics themselves

1) Basic network terms, OSI network model and TCP/IP protocol stack.
2) Upper level protocols.
3) Protocols of lower levels (transport, network and channel).
4) Network devices and types of cables used.
5) The concept of IP addressing, subnet masks and their calculation.
6) The concept of VLAN, Trunk and the VTP and DTP protocols.
7) Spanning Tree Protocol: STP.
8) Channel aggregation protocol: Etherchannel.
9) Routing: static and dynamic using the example of RIP, OSPF and EIGRP.
10) Network address translation: NAT and PAT.
11) First hop reservation protocols: FHRP.
12) Computer network security and virtual private networks: VPN.
13) Global networks and protocols used: PPP, HDLC, Frame Relay.
14) Introduction to IPv6, configuration and routing.
15) Network management and network monitoring.

P.S. Perhaps over time the list will be expanded.

So let's start with some basic networking terms.

What is a network? It is a collection of devices and systems that are connected to each other (logically or physically) and communicate with each other. This includes servers, computers, phones, routers, and so on. The size of this network can reach the size of the Internet, or it can consist of just two devices connected by a cable. To avoid any confusion, let’s divide the network components into groups:

1) End nodes: Devices that transmit and/or receive any data. These could be computers, phones, servers, some kind of terminals or thin clients, TVs.

2) Intermediate devices: These are devices that connect end nodes to each other. This includes switches, hubs, modems, routers, and Wi-Fi access points.

3) Network environments: These are the environments in which direct data transfer occurs. This includes cables, network cards, various types of connectors, and airborne transmission media. If it is a copper cable, then data transmission is carried out using electrical signals. In fiber optic cables, using light pulses. Well wireless devices, using radio waves.

Let's see it all in the picture:

On this moment you just need to understand the difference. The detailed differences will be discussed later.

Now, in my opinion, the main question is: What do we use networks for? There are many answers to this question, but I will highlight the most popular ones that are used in everyday life:

1) Applications: Using applications, we send various data between devices and open access to shared resources. These can be either console applications or GUI applications.

2) Network resources: This network printers, which, for example, are used in the office or network cameras that are viewed by security while in a remote area.

3) Storage: Using a server or workstation connected to the network, storage is created that is accessible to others. Many people post their files, videos, pictures there and open general access to them for other users. An example that comes to mind is google drive, Yandex disk and similar services.

4) Backup: Often, large companies use a central server where all computers copy important files for backup. This is necessary for subsequent data recovery if the original is deleted or damaged. There are a huge number of copying methods: with preliminary compression, encoding, and so on.

5) VoIP: Telephony using IP protocol. It is now used everywhere, as it is simpler and cheaper traditional telephony and every year it displaces it.

Of the entire list, most often many worked with applications. Therefore, we will analyze them in more detail. I will carefully select only those applications that are somehow connected to the network. Therefore, I don’t take applications like a calculator or notepad into account.

1) Loaders. This file managers, working via FTP, TFTP protocol. A trivial example is downloading a movie, music, pictures from file hosting services or other sources. This category also includes backups that the server automatically makes every night. That is, they are built-in or third party programs and utilities that perform copying and downloading. This type applications do not require direct human intervention. It is enough to indicate the location where to save and the downloading will begin and end.

Download speed depends on bandwidth. For of this type applications this is not entirely critical. If, for example, a file takes 10 minutes to download, then it’s only a matter of time, and this will not affect the integrity of the file in any way. Difficulties can only arise when we need to do something in a couple of hours backup copy system, and due to a poor channel and, accordingly, low bandwidth, it takes several days. Below are descriptions of the most popular protocols in this group:

FTP It is a standard connection-oriented data transfer protocol. It works using the TCP protocol (this protocol will be discussed in detail later). The standard port number is 21. Most often used to upload a site to a web hosting and upload it. The most popular application, working on this protocol is Filezilla. This is what the application itself looks like:

TFTP- This is a simplified version of the FTP protocol that works without establishing a connection, using the UDP protocol. Used to load an image on diskless workstations. It is especially widely used by Cisco devices for the same image loading and backups.

Interactive applications. Applications that allow interactive exchange. For example, the “person-to-person” model. When two people, using interactive applications, communicate with each other or conduct general work. This includes: ICQ, Email, a forum where several experts help people with issues. Or the “man-machine” model. When a person communicates directly with a computer. This could be remote configuration of the database, configuration of a network device. Here, unlike bootloaders, constant human intervention is important. That is, at least one person acts as an initiator. Bandwidth is already more sensitive to latency than downloader applications. For example, when configuring a network device remotely, it will be difficult to configure it if the response from the command takes 30 seconds.

Real-time applications. Applications that allow you to transmit information in real time. This group includes IP telephony, streaming systems, and video conferencing. The most latency and bandwidth sensitive applications. Imagine that you are talking on the phone and what you say, the interlocutor will hear in 2 seconds and vice versa, you will hear from the interlocutor at the same interval. Such communication will also lead to the fact that voices will disappear and the conversation will be difficult to distinguish, and the video conference will turn into mush. On average, the delay should not exceed 300 ms. This category includes Skype, Lync, Viber (when we make a call).

Now let's talk about such an important thing as topology. It is divided into 2 large categories: physical And logical. It is very important to understand their difference. So, physical topology is what our network looks like. Where the nodes are located, what network intermediate devices are used and where they are located, what network cables are used, how they are routed and what port they are plugged into. Logical topology is which way packets will go in our physical topology. That is, physical is how we positioned the devices, and logical is which devices the packets will pass through.

Now let's look and analyze the types of topology:

1) Topology with a common bus (English Bus Topology)

One of the first physical topologies. The idea was that all devices were connected to one long cable and organized local network. Terminators were required at the ends of the cable. As a rule, this was a 50 ohm resistance, which was used to ensure that the signal was not reflected in the cable. Its only advantage was its ease of installation. From a performance point of view, it was extremely unstable. If there was a break somewhere in the cable, then the entire network remained paralyzed until the cable was replaced.

2) Ring Topology

In this topology, each device is connected to two neighboring ones. Thus creating a ring. The logic here is that at one end the computer only receives, and at the other it only sends. That is, a ring transmission is obtained and the next computer plays the role of a signal repeater. Due to this, the need for terminators disappeared. Accordingly, if the cable was damaged somewhere, the ring opened and the network became inoperable. To increase fault tolerance, a double ring is used, that is, each device receives two cables, not one. Accordingly, if one cable fails, the backup one remains operational.

3) Star topology

All devices are connected to the central node, which is already a repeater. Nowadays, this model is used in local networks, when several devices are connected to one switch, and it acts as an intermediary in transmission. Here the fault tolerance is much higher than in the previous two. If any cable breaks, only one device falls out of the network. Everyone else continues to work quietly. However, if the central link fails, the network will become inoperable.

4) Full-Mesh Topology

All devices are connected directly to each other. That is, from each to each. This model is, perhaps, the most fault-tolerant, since it does not depend on others. But building networks on such a model is difficult and expensive. Since in a network with at least 1000 computers, you will have to connect 1000 cables to each computer.

5) Partial-Mesh Topology

As a rule, there are several options. It is similar in structure to a fully connected topology. However, the connection is not built from each to each, but through additional nodes. That is, node A is connected directly only to node B, and node B is connected to both node A and node C. So, in order for node A to send a message to node C, it must first send to node B, and node B in turn will send this message to node C. In principle, routers operate on this topology. Let me give you an example from home network. When you go online from home, you do not have a direct cable to all nodes, and you send data to your provider, and he already knows where this data needs to be sent.

6) Mixed topology (English Hybrid Topology)

The most popular topology, which combines all the topologies above into itself. It is a tree structure that unites all topologies. One of the most fault-tolerant topologies, since if a break occurs at two sites, then only the connection between them will be paralyzed, and all other connected sites will work flawlessly. Today, this topology is used in all medium and large companies.

And the last thing left to sort out is network models. At the early stage of computers, networks did not have uniform standards. Each vendor used its own proprietary solutions that did not work with the technologies of other vendors. Of course, it was impossible to leave it like that and it was necessary to invent common decision. This task was undertaken by the International Organization for Standardization (ISO - International Organization for Standardization). They studied many models used at that time and as a result came up with OSI model, which was released in 1984. The only problem was that it took about 7 years to develop. While experts were arguing about how best to make it, other models were being modernized and gaining momentum. Currently, the OSI model is not used. It is used only as network training. My personal opinion is that every self-respecting administrator should know the OSI model like a multiplication table. Although it is not used in the form in which it is, the operating principles of all models are similar to it.

It consists of 7 levels and each level performs a specific role and task. Let's look at what each level does from bottom to top:

1) Physical Layer: determines the method of data transmission, what medium is used (transmission of electrical signals, light pulses or radio air), voltage level, and method of encoding binary signals.

2) Data Link Layer: it takes on the task of addressing within the local network, detects errors, and checks data integrity. If you have heard about MAC addresses and the Ethernet protocol, then they are located at this level.

3) Network Layer: this level takes care of combining network sections and choosing the optimal path (i.e. routing). Each network device must have a unique network address online. I think many have heard about the IPv4 and IPv6 protocols. These protocols operate at this level.

4) Transport Layer: This level takes on the function of transport. For example, when you download a file from the Internet, the file is sent in segments to your computer. It also introduces the concepts of ports, which are needed to indicate the destination to a specific service. The TCP (connection-oriented) and UDP (connectionless) protocols operate at this layer.

5) Session Layer: The role of this layer is to establish, manage, and terminate connections between two hosts. For example, when you open a page on a web server, you are not the only visitor on it. And in order to maintain sessions with all users, a session layer is needed.

6) Presentation Layer: It structures information in a readable form for the application layer. For example, many computers use an ASCII encoding table for output text information or jpeg format for graphic output.

7) Application Layer: This is probably the most understandable level for everyone. It is at this level that the applications we are familiar with work - e-mail, browsers using the HTTP protocol, FTP and the rest.

The most important thing to remember is that you cannot jump from level to level (For example, from application to channel, or from physical to transport). The entire path must go strictly from top to bottom and from bottom to top. Such processes are called encapsulation(from top to bottom) and deencapsulation(from lower to upper). It is also worth mentioning that at each level the information transmitted is called differently.

At the application, presentation and session levels, the transmitted information is designated as PDU (Protocol Data Units). In Russian they are also called data blocks, although in my circle they are simply called data).

Transport layer information is called segments. Although the concept of segments is applicable only to the TCP protocol. The UDP protocol uses the concept of a datagram. But, as a rule, people turn a blind eye to this difference.
At the network level they are called IP packets or simply packets.

And at the link level - frames. On the one hand, this is all terminology and it does not play an important role in how you call the transmitted data, but for the exam it is better to know these concepts. So, I’ll give you my favorite example, which helped me, in my time, understand the process of encapsulation and de-encapsulation:

1) Let’s imagine a situation where you are sitting at home at your computer, and in the next room you have your own local web server. And now you need to download a file from it. You type the address of your website page. Now you are using the HTTP protocol, which runs at the application layer. The data is packed and sent down to the next level.

2) The received data is sent to the presentation level. Here this data is structured and put into a format that can be read on the server. Packed up and lowered down.

3) At this level, a session is created between the computer and the server.

4) Since this is a web server and reliable connection establishment and control of received data is required, the TCP protocol is used. Here we indicate the port on which we will knock and the source port so that the server knows where to send the response. This is necessary so that the server understands that we want to get to the web server (standard is port 80), and not to mail server. We pack and move on.

5) Here we must specify which address to send the packet to. Accordingly, we indicate the destination address (let the server address be 192.168.1.2) and the source address (computer address 192.168.1.1). We turn it around and go down further.

6) The IP packet goes down and here the link layer comes into operation. It adds physical source and destination addresses, which will be discussed in detail in a subsequent article. Since we have a computer and a server in a local environment, the source address will be the computer’s MAC address, and the destination address will be the server’s MAC address (if the computer and server were on different networks, then addressing would work differently). If at the upper levels a header was added each time, then a trailer is also added here, which indicates the end of the frame and the readiness of all collected data for sending.

7) And the physical layer converts what is received into bits and, using electrical signals (if it is a twisted pair cable), sends it to the server.

The deencapsulation process is similar, but with the reverse sequence:

1) At the physical level they are accepted electrical signals and are converted into a understandable bit sequence for the link layer.

2) At the link layer, the destination MAC address is checked (whether it is addressed to it). If yes, then the frame is checked for integrity and absence of errors, if everything is fine and the data is intact, it transfers it to a higher level.

3) At the network level, the destination IP address is checked. And if it is correct, the data rises higher. There is no need to go into details now about why we have addressing at the link and network levels. This topic requires special attention, and I'll explain their differences in detail later. The main thing now is to understand how data is packed and unpacked.

4) On transport level The destination port (not the address) is checked. And by the port number, it becomes clear which application or service the data is addressed to. For us this is a web server and the port number is 80.

5) At this level, a session is established between the computer and the server.

6) The presentation layer sees how everything should be structured and makes the information readable.

7) And at this level, applications or services understand what needs to be done.

Much has been written about the OSI model. Although I tried to be as brief as possible and cover the most important things. In fact, a lot has been written in detail about this model on the Internet and in books, but for beginners and those preparing for CCNA, this is enough. There may be 2 questions in the exam for this model. This is the correct arrangement of layers and at what level a certain protocol operates.

As written above, the OSI model is not used nowadays. While this model was being developed, the TCP/IP protocol stack was becoming increasingly popular. It was much simpler and gained rapid popularity.
This is what the stack looks like:

As you can see, it differs from OSI and even changed the name of some levels. Essentially, its principle is the same as that of OSI. But only the three upper OSI layers: application, presentation and session are combined into one in TCP/IP, called application. The network layer has changed its name and is called the Internet. The transport one remained the same and with the same name. And the two lower OSI layers: channel and physical are combined in TCP/IP into one called the network access layer. The TCP/IP stack in some sources is also referred to as the DoD (Department of Defense) model. According to Wikipedia, it was developed by the US Department of Defense. I came across this question during the exam and before that I had never heard anything about her. Accordingly, the question: “What is the name network layer in the DoD model?”, threw me into a stupor. Therefore, it is useful to know this.

There were several other network models that lasted for some time. This was the IPX/SPX protocol stack. Used since the mid-80s and lasted until the late 90s, where it was superseded by TCP/IP. It was implemented by Novell and was an upgraded version of the Xerox Network Services protocol stack from Xerox. Used in local networks for a long time. The first time I saw IPX/SPX was in the game “Cossacks”. When choosing a network game, there were several stacks to choose from. And although this game was released somewhere in 2001, this indicated that IPX/SPX was still found on local networks.

Another stack worth mentioning is AppleTalk. As the name suggests, it was invented by Apple. It was created in the same year in which the OSI model was released, that is, in 1984. It didn't last long and Apple decided to use TCP/IP instead.

I also want to emphasize one important thing. Token Ring and FDDI are not network models! Token Ring is a link layer protocol, and FDDI is a data transfer standard that is based on the Token Ring protocol. This is not the most important information, since these concepts are not found now. But the main thing to remember is that these are not network models.

So the article on the first topic has come to an end. Although superficially, many concepts were considered. The most important ones will be discussed in more detail in the following articles. I hope now networks will no longer seem like something impossible and scary, and it will be easier to read smart books). If I forgot to mention something, have any additional questions, or if anyone has anything to add to this article, leave comments or ask in person. Thanks for reading. I will be preparing the next topic.

Add tags

About the fact that everything modern world is a giant virtual web known to, perhaps, every schoolchild. The times when the exchange of information was carried out on the principle of “hand to hand”, and the main data carrier was a stamped paper folder, are in the distant past, but now countless virtual highways connect all points of the planet into a single information system– computer data transmission network.

What is a computer network?

In a general sense, a computer data network is a communication system for various computer equipment (including PCs and user office equipment) necessary for automatic data exchange between end users, as well as remote control functional units And software this network.

There are a great many ways to classify computer networks (by architecture, type of transmission medium, network operating systems etc.), however, we will not delve into the jungle of the theory of network technologies: particularly inquisitive users will always be able to find this information in educational literature. Here we will limit ourselves to the simplest classification of networks depending on their length.

So, computer networks are divided on a territorial basis into local and global:

A global computer network is a data transmission network that covers the entire world (or individual large regions) and unites an unlimited number of unconnected subscribers.

A local computer network is a collection of PCs and network equipment connected by communication channels, designed to transmit data to a finite number of users. By the way, the term “local network” was assigned to the system at a time when the capabilities of the equipment did not allow organizing such communications for subscribers remote over long distances, but now local computer networks are used both for organizing local communications (within one building or organization), This covers entire cities, regions and even countries.

Types of computer networks

According to the method of organizing communication between subscribers, the topology of computer networks distinguishes the following local network schemes:

Where the network nodes are computers, office equipment and various network equipment.

More complex topologies (such as tree network, mesh network, etc.) are built by various connections of the three elementary types of local network.

Local network functions

We will not talk about the purpose of global networks and how the Internet benefits the world: the main functions of the World Wide Web are already well known to every user, and detailed description More than one book could be devoted to all the possibilities of the network.

At the same time, home networks are unfairly deprived of informational attention, and many users do not understand why they need a local network at all.

So, the main functions of a local network:

• - Optimization of the workflow. Thus, a home local network, organized, for example, in an office, provides all its employees with the opportunity to remotely exchange data, as well as share the use of all types of office equipment;
• - Communication. Of course, local networks will not be able to completely replace “Internet connection”, but in cases where it is necessary to organize your own communication channel, closed from external users (for example, a forum for corporate employees), local networks are simply irreplaceable;

• - Possibility of remote administration. Thus, a corporate local network allows one specialist to provide technical support several dozen different devices;
• - Saving. Agree, it is more logical to pay for an Internet connection once and provide all employees of the organization (user devices) with the opportunity to free access, than to pay for access to world wide web each employee (gadget) individually;
• - Games, security of data exchange, user comfort and much more.

Thus, the local network is very, very useful tool in any field of activity. In fact, it was local networks that replaced the well-known “ pigeon mail“both at any enterprise and between friends (after all, this is a much more functional alternative to tapping on the battery and “cactus” signals on the windowsill). And our lessons will help you not only create a local network from scratch with your own hands, but also solve much more complex issues of administering corporate networks and setting up different types of network equipment.