The topology of the computer network is determined. Design and calculation of the reliability and efficiency of the local computing network Reliability of the cable system of the Internet

The most important characteristic of computing networks is reliability. Improving reliability is based on the principle of fault prevention by reducing the intensity of failures and failures due to the use of electronic circuits and components with high and ultra-high degree of integration, reduction of interference levels, light operation modes, ensuring thermal modes of their work, as well as by improving the methods of assembling equipment .

Failure tolerance is such a property of a computing system that provides it as a logical machine the possibility of continuing actions, specified by the programAfter faults occur. The introduction of fault tolerance requires redundant hardware and software. Directions related to fault prevention and fault tolerance, basic reliability issues. On parallel computing systems is achieved both the highest performance and, in many cases, very high reliability. Available redundancy resources in parallel systems can be flexibly used both to increase productivity and improve reliability.

It should be remembered that the concept of reliability includes not only hardware, but also software. The main purpose of improving the reliability of systems is the integrity of the data stored in them.

Security is one of the main tasks solved by any normal computer network. Safety problem can be considered from different sides - malicious damage of data, confidentiality of information, unauthorized access, embezzlement, etc.

Provide information protection in the local network conditions is always easier than if there are a dozen autonomously working computers on the company. Almost at your disposal one tool is backup. For simplicity, let's call this process with reservation. Its essence consists in creating a full copy of the data updated regularly and as often as possible. For a personal computer, a floppy disk are served by a more or less safe medium. It is possible to use a streamer, but this is an additional cost of equipment.

Fig. 5.1. Data security tasks

It is easiest to ensure data protection from a variety of trouble in the case of a network with a dedicated file server. All the most important files are focused on the server, and save one car much easier than ten. The concentration of data facilitates both reservations, since they do not need to collect them all over the network.

Shielded lines allow you to improve the safety and reliability of the network. Shielded systems are much more resistant to external radio frequency fields.

They work, but not quite as I would like. For example, it is not very clear how to limit access to network disk, A printer at an accountant ceases to work every morning and there is a suspicion that the virus lives somewhere, because the computer has become unusually slowly working.

Familiar? You are not alone, these are classic signs of network service setup errors. This is quite corrected, we helped hundreds of times in solving such problems. Let's call it modernization of IT infrastructure, or improving reliability and security computer network .

Improving the reliability of the computer network - who is useful?

First of all, he needs an emergency to his company to the head. The result of a competently executed project is a significant improvement in the network and practically full elimination failure For this reason, the money spent on the modernization of the network in terms of improving IT infrastructure and increase the level of safety should be considered not costs, but investments that will definitely pay off.

Also, the network upgrade project is needed to normal users, since it allows them to focus on direct work, and not on solving IT problems.

How we carry out a network modernization project

We are ready to help you understand the problem, it is easy. Start with the fact that call us and ask to make an IT audit. He will show what causes daily problems, and how to get rid of them. We will make it for you either inexpensive or free.

In essence, IT audit is part of the network upgrade project. As part of the IT audit, we will not only examine the server and jobs, we will deal with the schemes for the inclusion of network equipment and telephony, but also develop a plan for the network upgrade project, we define the project budget both from the point of view of our work and the necessary equipment or software.

The next step is the actual implementation of the project to upgrade the network. Main works are manufactured on the server, since it is it that is a defining infrastructure component. Our task as part of the network modernization project is eliminated not so much manifestation as root problems. As a rule, they are reduced to approximately the same conceptual flaws of infrastructure:

a) servers and workstations work in the composition working Group, not a domain, as Microsoft recommends for networks with more than five computers. This leads to the problems of user authentication, it is impossible to effectively enter passwords and limit user rights, the inability to use security policies.

b) incorrectly configured network servicesIn particular DNS, and computers stop seeing each other or network resources. For the same reason, most often "inhibits the network" without any visible reasons.

c) on computers installed antivirusthat turns protection into a colander. You can work for years on a slow car without suspecting that 80% of its resources are busy to attack other computers or send spam. Well, be able to steal your passwords or transfer to an external server all you write. Unfortunately, it is quite real, reliable antivirus defense It is an important and necessary part of the project to upgrade any network.

These are the three most common causes of infrastructure problems, and each of them means the need for urgent elimination. It is necessary not to simply fix the problem, but also competently build a system to eliminate the very possibility of their appearance.

By the way, we try to use the phrase "Modernization of the information system" instead "Network Modernization"Since we try to watch wider network problems. In our opinion, the information system should be considered from various points of view, and a professional, developing a project to modernize the network, should take into account the following aspects of its work.

Information security of your company

Speaking about the information security of the company, we consider it very important not so much external protection against invasions over the Internet, how much to streamline the internal work of employees. Unfortunately, unknown hackers are applied the greatest damage of the company, but those people you know in person, but who may be offended by your decisions or consider information to be their property. The customer belonging manager, or an offended employee who coping accounting or management information "just in case" is the two most frequent cases of informational security.

Save data

Unfortunately, the preservation of the data very rarely appears on the list of attention of the head and even many IT professionals. It is believed that since space ships go with orbits, then it is almost impossible to prevent the breakage of the server. And the network modernization project often does not cover this part of the infrastructure.

Partly agree that it is not always possible to prevent an accident. But to make it so that the data always remains safe and maintained, and the company's work could be restored within one hours and two from the moment the server breakdowns can and you need to any self-respecting IT specialist. We consider our duty during the project to modernize the network to implement both hardware diagrams of information storage media and backup data by a special scheme that allows you to restore the data at the right time and ensure their safety for a long time. And if the administrator does not understand the meaning of the words mentioned above, then he, let's say gently, does not deserve confidence as a professional.

Long-term equipment work

Long-term operation of servers and workstations is directly related to what they are made and how. And we try to help choose such equipment that is bought for a long time and which does not require attention for many years. And within the framework of the project to modernize the network, it is very often to modernize that the server's disk subsystem is unfortunately, it is often forgotten about it. This is because the real service life hard disks Does not exceed 4 years, and after this time, they are subject to replacement on servers. This should be tracked as part of servicing servers and computers, because it is very important for the reliability of data storage.

Server and Computer Systems Maintenance

We should not forget that even a very correctly structured and reliable infrastructure requires competent and attentive maintenance. We believe that IT outsourcing in terms of infrastructure service is a logical continuation of project work. There are a number of companies that have their own IT professionals, but the tasks of maintaining the server systems entrusted to us. This practice shows high efficiency - the company pays only for the support of servers, taking low-level tasks for themselves. We are responsible for ensuring that security and backup policies are followed in order to carry out regulatory work, monitor server systems.

The relevance of IT solutions

The world is constantly changing. The IT world is changing twice. And technology is born and dying faster than we would like to spend money on their update. Therefore, conducting a network modernization project, we consider it necessary to introduce not only the newest, but also the most reliable and justified solutions. Not always what everyone talks about is a panacea or solving your problem. Often, everything is not exactly as described. Virtualization, cloud computing are used by thousands of companies, but not always the introduction of some technologies is economically justified. And vice versa - the correctly selected and competent network modernization project and reasonable choice Software gives new opportunities in work, saves time and money.

Paid Windows or free Linux? MS SharePoint or "Bitrix: Corporate Portal"? IP telephony or classic? Each product has its advantages and its own scope of application.

What do your company need? How to perform a network upgrade project or introducing a new service so as not to interrupt the company's job? How to make the introduction to be successful, and employees got the best tools for work? Call us, let's figure it out.

Send your good work in the knowledge base is simple. Use the form below

Students, graduate students, young scientists who use the knowledge base in their studies and work will be very grateful to you.

Posted by http://www.allbest.ru/

Introduction

local computing network

To date, there are more than 130 million computers in the world, and more than 80% of them are combined into various information and computing networks, from small local networks in offices, to global networks type of Internet.

The experience of operating networks shows that about 80% of the entire information sent over the network will be closed within a single office. Therefore, the special attention of developers began to attract the so-called local computing networks.

The local network is a set of computers, peripheral devices (printers, etc.) and switching devices connected by cables.

Local computing networks differ from other networks by the fact that they are usually limited to a moderate geographic area (one room, one building, one district).

A lot depends on the quality and thoughtful performance of the initial stage of implementation of the LAN - from the pre-project examination of the system of document management of that enterprise or that organization, where it is planned to establish a computing network. It is here that such major network indicators are laid such as its reliability, spectrum of functionality, service life, continuous performance, service technology, working and maximum network utilization, network security, and other characteristics.

The worldwide tendency to combine computers on the network is due to the number of important reasons, such as accelerating the transfer of informational messages, the ability to quickly exchange information between users, receiving and transmitting messages, without departing from the workplace, the possibility of instant receipt of any information from any point of the globe, as well as Exchange information between computers of various manufacturers operating under different software.

Such huge potential capabilities that the computing network carries, and that new potential lift, which at the same time experiences the information complex, as well as a significant acceleration of the production process, do not give us the right not to take it into developing and not apply them in practice.

1. The purpose of the work.

The aim of the work is to obtain the skills of developing the structure of local computing networks, the calculation of the main indicators that determine the operation of the network.

2. Theoretical part

2.1. Basic goals of creating local computing network (LAN).

The constant need to optimize the distribution of resources (primarily information) periodically puts us before the need to develop a fundamental solution to the issue of the organization IVS (information and computing network) on the basis of an already existing computer fleet and a software package that meets modern scientific and technical requirements, taking into account the increasing needs and the possibility of further gradual development of the network in connection with the emergence of new technical and software solutions.

You can briefly allocate the main advantages of the application of the LAN:

Separation of resources

The separation of resources allows economically to use resources,

for example, manage peripheral devices, such as laser printing devices, from all attached workstations.

Data separation.

Data separation provides the ability to access and manage databases from peripheral jobs in need of information.

Separation of software

Software separation provides the possibility of simultaneously use of centralized, previously installed software.

Processor resource separation

When separating processor resources, it is possible to use computational power to process data by other systems that are included in the network.

Opendetails and Terminology

The local computing network (LAN) is a high-speed communication line of hardware data processing in a limited area. LAN can combine personal computers, terminals, minivars and universal computing machines, printing devices, voice information processing systems and other devices

Network devices (SU) - specialized devices intended for collecting, processing, converting and storing information received from others network devices, workstations, servers, etc.

The main component of the local computing network is the Local Computing Network Workstation (RSLVS), i.e., computer, the hardware capabilities of which allow you to exchange information with other computers.

Local computing network is a complex technical systemwhich is a combination of hardware and software, since a simple connection of devices, however, does not mean the possibility of their collaboration. For efficient communication of various systems, the corresponding software. One of the main functions of the LAN operational support is to maintain such a connection.

The rules of the seeds - how the system produces a survey and must be interviewed - called protocols.

Systems are called similar if they use the same protocols. When using different protocols, they can also work due to each other using the software that makes the mutual transformation of the protocols, the LAN can be used to communicate not only the PC. They can associate video systems, systems telephone communication, Production equipment and almost everything that requires high-speed data exchange. Several local computing networks can be combined through local and remote communications in firewall mode.

Personal computers are combined into a network mainly to share programs and data files, messaging messages (mode email) and to share resources (printing devices, modems and hardware and programmatic communications). In this case, personal computers refer to the workstations of the local computer network.

Modern technology of local computing networks allows you to use different types Cables in the same network, as well as unhindered to connect various LAN equipment in one network, such as Ethernet, ArchNet, and Token-Ring.

Perdachas solved when creating a LAN

Creating a LAN, the developer is faced with a problem: with the known data on the purpose, the list of LAN functions and the basic requirements for a complex of technical and software LANs to build a network, that is, to solve the following tasks:

determine the LAN architecture: Select the types of the LAN component;

make an assessment of performance indicators of the LAN;

determine the value of the LAN.

It should take into account the rules for the connection of the LAN components based on network standardization, and their limitations specified by manufacturers of the LAN component.

The LAN configuration for ACS significantly depends on the characteristics of a particular application area. These features are reduced to the types of transmitted information (data, speech, graphics), the spatial location of subscriber systems, information flow intensities, permissible information delays in transmission between sources and recipients, data processing volumes in sources and consumers, subscriber station characteristics, external climatic, electromagnetic factors, ergonomic requirements, requirements for reliability, the cost of LAN, etc.

Determining the network topology

Consider the topology options and the composition of the component of the local computing network.

The topology of the network is determined by the method of connecting its nodes of communication channels. In practice, 4 basic topologies are used:

star-shaped (Fig. 1, A, 1, b);

ring (Fig. 2);

tire (Fig. 3);

tree or hierarchical (Fig. 4).

AK - Active PC Hub - Passive Concentrator Fig. 4. Hierarchical network with concentrators.

The selected network topology must correspond to the geographical location of the LAN network, the requirements set for the network characteristics listed in Table. one.

Table 1. Comparative data on the characteristics of the LAN.

Selection of the type of communication tools. Twisted para

The cheapest cable connection is a twisted pair "twisted pair". It allows you to transfer information at a speed of up to 10 Mbit / s, it is easily increasing, however, is oxo-proof. The cable length cannot exceed 1000 m at a transmission rate of 1 Mbps are the low price and trouble-free installation, the shielded vita pair is often used to increase the noise immunity, ie, vitua pair, placed in the shielding shell, like the coaxial cable screen. It increases the cost of twisted pair and brings its price to the price of a coaxial cable,

Coaxial cable

The coaxial cable has an average price, well-protein and is used for long distance communications (several kilometers). Information transmission rate from 1 to 10 Mbps, and in some cases it can reach 50 Mbps C-coaxial cable is used for basic and broadband information transmission,

Broadband coaxial cable

Broadband coaxial cable is immune to interference, it is easily increasing, but its price is high. The speed transmission rate is 500 Mbps, when transmitting information in the baseband frequency band, the amplifier is required for a distance of more than 1.5 km, or the so-called repeater (repeater), so the total distance during information transfers increases to 10 km. For computing networks with a tide topology or a tree, the coaxial cable must have at the end of the agreement resistor (terminator).

Ethernet cable

Ethemet cable is also a coaxial cable with a wave resistance of 50 ohms. It is called even fat Ethernet (Thick) or a yellow cable (Yellow Cable).

Due to noise immunity is an expensive alternative to ordinary coaxial cables. The maximum available distance without the repeater does not exceed 500 m, and the total distance of the Ethernet network is about 3000 m. Ethernet cable, due to its main topology, uses only one load resistor at the end.

Cheapernet - Cable

Cheaper than Ethernet Cable is a Cheapernet cable connection or, as it is often called, thin (THIN) Ethernet. It is also a 50-ohm coaxial cable with a speed transfer rate of ten million bits. When connecting the segments of the Cheapernet cable, repeaters are also required. Computing networks with Cheapernet cable have a small cost and minimum costs When building. Network circuit boards are made using widely used small-sized bayonet connectors (CP-50). Additional shielding is not required. The cable joins the PC using tee connectors (T-Connectors). The distance between two workstations without repeaters can make up a maximum of 300 m, and the total distance for the network on the Cheapernet cable is about 1000 m. The Chepernet transceiver is located on the network board and both for electroplating between adapters and to enhance the external signal.

Fiber optic lines

The most expensive are optical solutions, also called fiberglass cable. The speed of dissemination of information on them reaches several gagabit per second. Permissible removal of more than 50 km. External impact of interference is practically absent. At the moment it is the most expensive connection for the LAN. Applied where electromagnetic fields of interference occurs or information is required for very long distances without the use of repeaters. They have via contradictory properties, since the techniques of branches in fiber optic cables are very complex. Optic strokes are combined into a LAN with a star-shaped connection.

Selecting a type of construction setand according to the information transfer method

Local Network Token Ring

This standard is developed by IBM, as a transmitted environment, unshielded or shielded twisted pair (UPT or SPT) or fiber is used. Data transfer rate 4 Mbps or 16 MBIT / s. As a method for managing the access station to the transmitting environment, a method is used - a marker ring (Token Ring). The main provisions of this method:

Devices are connected to the network on the topology ring;

All devices connected to the network can transmit data only by receiving a transmission permission (marker);

at any time, only one station in the network has such a right.

Online can connect computers on the topology star or ring.

Local network Arcnet

ARKNET (Attached Resource Computer Network) is a simple, inexpensive, reliable and fairly flexible architecture of the local network. Developed by Datapoint Corporation in 1977. Subsequently, an SMC (Standard MicroSistem Corporation) has acquired a license for ArcNet, which has become the main developer and manufacturer of equipment for ArcNet networks. As a transmitted medium, twisted steam, coaxial cable (RG-62) with a wave resistance of 93 Ohm and fiber-optic cable, data transfer rate - 2.5 Mbps. When connecting devices in ArcNet, the topologies of the bus and the star are used. Method of stations access control method to the transmitting medium - marker tire (TOKEN BUS). This method provides for the following rules:

At any time, only one station in the network has such a right;

Basic principles of work

The transfer of each byte in ArcNet is performed by a special ISU send (information Symbol Unit - an information transfer unit) consisting of three service start / stop bits and eight data bits. At the beginning of each package, the initial separator Alert Burst is transmitted, which consists of six service bits. The initial separator performs the package preamble functions.

On the ARCnet network you can use two topologies: Star and Tire,

Local network Ethernet

The Ethernet specification at the end of the seventies was offered Xerox Corporation. Later, Digital Equipment Corporation (DEC) and Intel Corporation also joined this project. In 1982, the specification on Ethernet version 2.0 was published. Based on Ethernet Institute IEEE has been developed standard IEEE. 802.3. Differences between them are minor.

Basic principles of work:

On the logical level in Ethernet uses the tire topology;

All devices connected to the network are equal, i.e. any station can start transmission at any time (if the transmission medium is free);

Data transmitted by one station is available to all network stations.

ChooseoP network operating system

A wide variety of types of computers used in computing networks entails a variety of operating systems: for workstations, for the system networks and servers of the enterprise level of the enterprise as a whole. These may be presented with various performance requirements and functionality, it is desirable that they have a compatibility property that would allow to ensure the joint work of various OS. Network OS can be divided into two groups: the scale of the department and the scale of the enterprise. OS for departments or working groups provide a set of network services, including dividing files, applications and printers. They should also provide fault tolerance properties, such as working with RAID arrays, maintain cluster architecture. The networks of departments are usually simpler in the installation and control compared to the network OS enterprise, they have less functional properties, they less protect the data and have weaker interaction capabilities with other types of networks, as well as worse performance. Network operating system The scale of the enterprise must first have the main properties of any corporate products, including:

scalability, that is, the ability to work equally well in a wide range of various quantitative characteristics of the network,

compatibility with other products, that is, the ability to work in a complex heterogeneous interset environment in Plug-and-Play mode.

Corporate network OS must support more complex services. Like the network operating system, the network operating system should allow users to share files, applications and printers, and do it for more users and data volume and with higher performance. In addition, the network-scale enterprise provides the ability to connect heterogeneous systems - both workstations and servers. For example, even if the OS is working on the Intel platform, it must support UNIX workstations running on RISC platforms. Similarly, the server OS operating on the RISC computer must support DOS, Windows and OS / 2. The network scales of the enterprise must support multiple protocol stacks (such as Tsryar, IPX / SPX, NETBIOS, DECNET and OSI), providing easy access to remote resources, convenient service management procedures, including agents for network management systems.

An important element of the network operating system of the enterprise is the centralized reference service in which data on users and the shared network resources are stored. Such a service, also called the directories, provides a single logic input of the user to the network and provides it with convenient means of viewing all resources available to it. The administrator, if there is a centralized reference service in the network, is delivered from the need to start a repeated list of users on each server, which means the large number of routine work and from potential errors in determining the composition of users and their rights on each server. An important property of the reference service is its scalability provided by the distribution of database of users and resources.

Networks such as Banyan Vines, Novell Netware 4.x, IBM Lan Server, Sun NFS, Microsoft Lan Manager and Windows NT Server, can serve as an enterprise operating system, while NetWare 3.x, Personal Ware, Artisoft Lantastic is more suitable for small working groups.

The criteria for the choice of the enterprise scale are the following characteristics:

Organic support of the multi-server network;

High efficiency of file operations;

The possibility of effective integration with other OS;

Availability of centralized scalable reference service;

Good development prospects;

Effective work of remote users;

Diverse Services: File service, Print service, Data Security and Failure tolerance, Data Archiving, Messaging Service, Diverse Databases and Others;

A variety of transport protocols: TCP / IP, IPX / SPX, NetBIOS, AppleTalk;

Support for diverse operating systems of end users: DOS, UNIX, OS / 2, MAC;

Support for network equipment Ethernet standards, token Ring, FDDI, ArcNet;

The presence of popular application interfaces and call mechanisms of remote RPC procedures;

The ability to interact with the network control and management system, support for SNMP network management standards.

Of course, none of the existing network OS is in full compliance with the following requirements, so the selection of the network OS is usually carried out taking into account the production situation and experience. The table shows the main characteristics of popular and affordable networks.

Determination of the reliability of the work of the LAN. 2.4.1. Preliability Reliability LAN

In general, reliability is the property of a technical device or product to perform its functions within allowable deviations for a certain period of time.

The reliability of the product is laid at the design stage and significantly depends on such criteria as the choice of technical and technological specification, the compliance of the world-class adopted design solutions. The reliability of the LAN also affects the literacy of the staff at all levels of networking, the conditions for transportation, storage, installation, setup and running each network node, compliance with the rules of equipment.

When calculating and evaluating the reliability of the computing network, the following terms and definitions will be used:

Performance is the condition of the product in which it can perform its functions within the established requirements.

Failure - an event in which the performance of the product is disturbed.

Malfunction is the condition of the product at which it does not correspond to at least one requirement of technical documentation.

Working is the duration of the product in hours or other units of time.

Working on the failure, or the average time of trouble-free operation - the average value of the work of the repaired product between failures.

The probability of trouble-free operation is the likelihood that this period of time does not arise failure.

Failure intensity - the probability of failure of an unrestregnosed product per unit time after this moment time.

Undetyability - product property to maintain performance for some occasions.

Durability - product property to maintain performance before the limit state with maintenance and repair interruptions.

Resource - product work before the limit state specified in the technical documentation.

Service life - the calendar duration of the work of the product to the limit state specified in the technical documentation.

Maintainability - product availability for its service

and repair.

Reliability is a comprehensive property that includes such properties as:

performance;

persistence;

maintainability;

durability.

The main property described by quantitative characteristics is operational.

Loss of workability - failure. Refusals of the electrical product can mean not only electrical or mechanical damage, but also the care of its parameters for the permissible limits. In this regard, the failures may be sudden and gradual.

The emergence of sudden failures in the device are random events. These failures can be independent when the failure of one element in the device is regardless of other elements, and dependent when the failure of one element is caused by the refusal of others. The separation of failures on sudden and gradual is conditional, since sudden failures may be caused by the development of gradual failures.

Main quantitative characteristics of reliability (performance):

probability of trouble-free work during T: P (T);

the probability of failure for T: Q (T) \u003d 1 - P (T);

the intensity of failures X (T) - indicates the average number of failures arising over the unit of operation time of the product;

the average time of product operations to failure (value, reverse failure intensity).

The real values \u200b\u200bof the specified characteristics are obtained by the results of reliability tests. In the calculation of time to failure / is considered a random value, therefore, the apparatus of probability theory is used.

Properties (axioms):

P (0) \u003d 1 (Operation of operational products);

lim t _\u003e 00 p (t) \u003d o (performance cannot be saved unlimited time);

dP (T) / DT<0 (в случае если после отказа изделие не восстанавливается).

During the service life of the technical device, three periods can be distinguished, the intensity of the failures in which varies in different ways. The dependence of the intensity of failures is shown in Fig. 5.

Fig.5. Typical Changes Change X (T) during the life of the product.

I - DX (T) / DT Training Stage<0

II - stage of normal operation X (T) -Const

III - step of aging DX (T) / DT\u003e 0

In the first period, called the period of work, there is a detection of constructive, technological, assembly and other defects, so the intensity of failures can rise at the beginning of the period, falling when approaching the period of normal operation.

The period of normal operation is characterized by sudden failures of constant intensity, which increases by the wear period.

During the period of wear, the failure intensity increases over time as the product wear.

Obviously, the main period should be the period of normal operation, and other periods are periods of entry and exit from this period.

Axiom 3 is valid for non-standard elements (microcircuits, radio elements, etc.). The process of operation of the recoverable systems and products differs from the same process for non-stipped in the way that, along with the flow of failures of the product elements, there are stages of repairing the refused elements, i.e. There is a stream of recovery of elements. For recoverable systems, the third property is not performed: DP (T) / DT<0. За период времени At могут отказать два элемента системы, а быть восстановленными - три аналогичных элемента, а значит производная dP(t)/dt>0.

When configuring computing networks, it operates with such a concept as the average time of operation for the failure of a network element (TN).

For example, if 100 products were tested during the year and 10 of them failed, TN will be equal to 10 years. Those. It is assumed that after 10 years all products will fail.

Quantitative characteristic for mathematical determination of reliability is the intensity of the device failures per unit of time, which is usually measured by the number of failures per hour and is indicated by the X icon.

The average time of operational failure and the average recovery time of working capacity is related through the coefficient of readiness kg, which is expressed in the likelihood that the computing network will be in a working condition:

Thus, the coefficient of readiness kg of the entire network will be determined as a product of the private coefficient of readiness KRI. It should be noted that the network is considered reliable at kg\u003e 0.97.

An example of calculating reliabilityand local computing network

The local computing network usually includes a set of user workstations, a network administrator workstation (one of the user stations can be used), the server kernel (a set of hardware server platforms with server programs: file server, www server, bd server, mail server etc.), communication equipment (routers, switches, hubs) and structured cable system (cable equipment).

The reliability calculation of the LAN is starting with the formation of the concept of failure of this network. For this, management functions are analyzed, the execution of which in the enterprise is carried out using this LAN. These features are selected, the violation of which is unacceptable, and the LAN equipment is defined, involved in their execution. For example: Certainly during the working day, the ability to call / write information from the database, as well as access to the Internet.

For the combination of such functions according to the structural electrical circuit, the LAN equipment is determined, the refusal of which directly disrupts at least one of specified functionsand a logical scheme for calculating reliability is drawn up.

At the same time, the quantities and conditions for the operation of repair and recovery brigades are taken into account. The following conditions are usually accepted:

Restoration limited - i.e. At any time, it cannot be recovered more than one failed element, because There is one repair brigade;

the average recovery time of the failed element is established or based on the permissible breaks in the work of the LAN, or from the technical capabilities of delivery and inclusion in the operation of this element.

Within the framework of the above approach to the calculation of the scheme for calculating reliability, as a rule, it can be reduced to a sequential-parallel scheme.

We establish a refusal of equipment included in the kernel of the network: servers, switches or cable equipment as a criterion to refuse LAN. We believe that the failure of the workstations of users does not lead to the failure of the LAN, and since the simultaneous failure of all workstations is an unlikely event, the network during individual failures of workstations continues to function.

Fig.6. Scheme of LAN elements for calculating total reliability.

Let's take it that the local network Includes two servers (one provides an output to the Internet), two switches and five cable fragments belonging to the network kernel. The intensity of failures and recovery for them is shown below.

In this way,

1) the intensity of the failure of the entire network L is 6.5 * 10-5 1 / h,

2) the average time of operation on the failure of the entire TN network is approximately 15.4 thousand.

3) The average time recovery time is 30 hours.

The calculated values \u200b\u200bof the corresponding readiness are presented in Table. four:

The coefficient of readiness of the entire network is

Calculation of the effectiveness of the work of the LAN

To determine the parameters of the network functioning, the choice and justification of control points. For these selected points, the information is collected and the calculation of the parameters:

request processing time is the calculation of the time interval between the formation of the query and receiving an answer to it performed for the selected basic services.

the reaction time in the loaded and unloaded network is the calculation of the performance indicator unloaded and unloaded network.

the frame transfer delay time is to calculate the delay time of the channel level of the selected main network segments.

definition real bandwidth - Definition of real bandwidth for routes of selected main network nodes.

analytical calculation of reliability indicators - Analytical assessment of the possible intensity of failures and average time for failure.

the readiness coefficient is an analytical calculation of the degree of readiness (average recovery time) LAN.

Suppose that the network between two users is organized according to the scheme shown in Fig. 7.

Procedure for performing work

To perform work it is necessary:

a) repeat the safety regulations when working with computing equipment;

b) examine lecture materials on courses "", as well as theoretical part of these methodological instructions;

c) choose a semi-power enterprise or organization and explore it existing system Document management from the point of view of automation. Suggest a new document management system based on the application of computing networks, assess the advantages and disadvantages of the existing and proposed systems (speed, cost, topology, changes in the wage fund, etc.);

d) Calculate numeric indicators of the new document management system: the reliability of the network, the time of operation for failure, the readiness ratio, the delivery time of the message to the addressee, the time of receipt of the receipt of the message delivery;

d) in accordance with the requirements given in section 5, issue a report on laboratory work;

g) protect the laboratory work, demonstrating the teacher:

1) a report on laboratory work;

2) understanding the basic principles of the organization of the local computing network;

3) Theoretical knowledge of the quantitative parameters of the computer network.

When preparing for protection for self-test, it is recommended to respond to the control questions shown in section 5.

4. Report requirements

Laboratory report report must contain:

a) title sheet;

b) the assignment condition;

c) justification for the development of LAN and calculations on the proposed network topology;

d) comments and conclusions on the work done.

Bibliography

1.Guseva A.I. Work in local networks NetWare 3.12-4.1: textbook. - M.: Dialog Mafi, 1996. - 288 p.

2.Orin Distributed computing systems:. - M.: Radio and Communication, 1984. - 296 p.

4.Frolov A.V., Frolov G.V. Local networks of personal computers. Using IPX, SPX, NetBIOS.- m.: "Dialog Mafi", 1993. - 160 s.

Posted on Allbest.ru.

...

Similar documents

Local computing network, switching and communication lines, providing network user data transfer. Channel level of the OSI model. Computer location scheme. Calculation of the total cable length. Software I. hardware local network.

course work, added 06/28/2014


Methods for communicating disparate computers to the network. Basic principles of the organization of a local computing network (LAN). Development and design of a local computing network in the enterprise. Description of the selected topology, technology, standard and equipment.

thesis, added 06/19/2013


The objectives of the informatization of school No. 15 of the Volga region. Design and organization of the school network. Structure and basic functions of a local computer network. Characteristics of software and hardware, construction mechanisms and features of the administration of the LAN.

thesis, added 05/20/2013


Justification of the modernization of the local computing network (LAN) of the enterprise. Equipment and software LAN. Selection of network topology, cable and switcher. Implementation I. wi-Fi Setup - Access points. Ensuring the reliability and security of the network.

thesis, added 12/21/2016


Creating a local computer network, its topology, cable system, technology, hardware and software, minimum requirements to the server. Physical construction of the local network and the organization of the Internet access, the calculation of the cable system.

course work, added 05.05.2010


Computer Local Computer Network: Design on two floors, interaction about 30 cars. The distance between the machines and switches is at least 20 meters, the number of switches - within the project. Logical and physical network topology.

laboratory work, added 09/27/2010


Main types of communication lines. Local computing networks (LAN) as a distributed data processing system, features coverage of the territory, cost. Analysis of the possibilities and relevance of the use of network equipment in the construction of modern LANs.

thesis, added 16.06.2012


Calculations of the parameters of the designed local computing network. Total cable length. Distribution of IP addresses for the designed network. Specification of equipment and consumables. Select the operating system and application software.

course work, added 01.11.2014


Overview of the design methods of the local computing network for training premises of one of the college buildings according to the Ethernet standard using the "twisted pair" cable and the "fine coaxial" in all parameters using 10Base-T and 10Base standards.

coursework, added 03/24/2011


The main stages of maintenance and modernization of the local network of the enterprise. Type of automated activity in the enterprise. Select the topology of the local computer network. Hardware and software. Characteristics of the seven-level OSI model.

Content
Introduction
Protection against errors in the transmitted information in the COP
Distribution of resources in networks
Protection and emergency recovery of information in the COP
Conclusion
Bibliography

Introduction
For many years ago, personal computers (PCs) were used independently small small islands of computing power, inhabiting tables in homes and offices. And the very fact that on each PC was often different from other version of any operating system or an application, perceived no more than annoying trouble.
Years and network technology were held for personal computers, and users began to understand that they could work together. Employment of personal computers with the ability to interact with each other discovered tremendous opportunities for cooperation and joint activities. Today, computer networks are vital for the functioning of all types of business and are found even at home, combining several PCs. With competent tools and configuration, computer networks can be very fast and reliable in operation.
However, networks may fail and when malfunctions are required, it is necessary to perform decisive actions to detect and correct the problem. And if we take into account that in addition to cables, hubs, routers, switches and other network devices, many computer networks may include hundreds and even thousands of PCs, it becomes clear that it is required to effectively eliminate faults more than just replacing personal computers and other networks. devices.

Protection against errors in the transmitted information in the COP
The reliability of the network is related to the ability to transmit reliably (without errors) user data from one OOD (terminal data terminal) to another OOD. It includes the ability to recovery after errors or data loss on the network, including channel failures, ODA, ACD (data channel termination equipment) or OKD (data switching equipment). Reliability is also associated with the maintenance of the system, which includes daily testing, preventive maintenance, for example, replacing the refused or permissible components; Diagnosing malfunction in malfunctions. In case of a problem with any component, a network diagnostic system can easily detect an error, localize a malfunction and, possibly disable this component from the network. The appearance of errors in transmitting information is explained by either foreign signals, always present in channels, or interference caused by external sources and atmospheric phenomena, or other reasons. In telephony distortion, it is considered to change the current form in the receiving unit, and in the telegraph - a change in the duration of the received parcels of the current compared to the transmitted parcels.
"Violations" or error can be widely classified as random, impulse and mixed.
Random errors occur by chance in the blocks of received data. Most of the channels with real carriers (as well as satellite channels) are subject to random errors.
Channels with pulse errors demonstrate a status free from errors, most of the time, but sometimes group or one-time errors appear. The object of such errors are radio signals, as well as cables and wires, such as telephone channels from twisted wired steam.
To improve the accuracy and quality of communication systems, group methods of protection against errors, excess coding and system with feedback. In practice, a combination combination of these methods is often used. For group methods of protection from errors, a method already used in the telegraph, known as the principle of Verdan, has long been used in telegraph: all information (or individual code combinations) is transmitted several times, usually no even number of times (minimum three times). The received information is remembered by a special device and compared. The judgment about the correctness of the transfer is made by the coincidence of most of the information received by the methods "two of three", "three out of five" and so on.
Another method, which also does not require the transcoding of information, implies information transmission by blocks consisting of several code combinations. At the end of each block, information is sent containing the quantitative characteristics of the transmitted block, for example, the number of units or zeros in the block. At the receiving end, these characteristics are reincarnated again, compared with the communication channel transmitted via the communication channel, and if they match, the block is considered accepted correctly. If the quantitative characteristics are insecured, the error signal is sent to the transmitting side.
Among the protection methods from errors, no propagation was made of noise-resistant encoding, allowing to obtain higher qualitative performance of communication systems. Its main purpose is to take all possible measures to ensure that the probability of distortion of information is enough, despite the presence of interference or failures in the network. Noise-resistant encoding involves the development of corrective (noise-resistant) codes that detect and correct a certain kind of error, as well as the construction and implementation of coding and decoding devices.
When transmitting information, depending on the number of codes, codes can be two-position and multi-position. According to the degree of noise immunity, two-position codes are divided into ordinary and noise-resistant.
Two-position ordinary codes are used to transmit all possible elements Code combinations and are uniform when the length of all code combinations is the same, for example, five-element telegraph code, and uneven when code combinations consist of a different number of items, such as Morse code.
In noise-resistant codes, except for information elements, one or more additional elements that are verification and employees to achieve a higher quality data transmission are always contained. The presence of redundant information in codes allows you to detect and correct (or only detect) errors.
The choice of corrective codes to a certain extent depends on the requirements for the reliability of the transfer. For its correct choice, it is necessary to have statistical data on the patterns of errors, their character, number and distribution over time. For example, a corrective code, correcting single errors, can be effective only if the errors are statistically independent, and the likelihood of their appearance does not exceed some. This code is absolutely not suitable if the errors appear by groups (packs). Recurrent codes that correct group errors may also be ineffective if the number of errors during transmission will be greater than the permissible norm.
Developed various corrective codes are divided into continuous and block. In continuous, or recurrent, codes
Control elements are located between the information. In blocks
codes information is encoded, transmitted and decoded by individual groups (blocks) of equal length. Block codes are separable (all information and control elements are placed on strictly defined positions) and inseparable (elements of the code combination do not have a clear division into redundant and informational). Inseparable include code with a constant number of zeros and units.
Sliding codes consist of systematic and non-systematic. In systematic codes, test symbols are formed using various linear combinations. Systematic codes - the most extensive and most applied group of corrective codes. They include such codes as Hamming code, cyclic codes, Bowza Chowudhuri codes and others. Large computing systems (AMDAL, IBM, Burroughs, ICL) use a very complex method of checking errors when transmitting over lines of communication between machines. The PEVM typically uses a more simple error verification technique. One of the simplest forms of error check is the so-called echoplex. In accordance with this technique, each character sent by the PEVM on the duplex line of the remote subscriber returns back to the PEVM in the form of echo. If the PC takes the same symbol as sent, it is understood that the transmission of the symbol passed correctly. If not, it means that an error occurred when transmitted and re-transmitting the same symbol. The echoplex is used in bidirectional duplex communication channels.
Another commonly used in practice (and relatively simple) method is to control the parity. Its essence is that each code combination is added to one discharge in which a unit is recorded if the number of units in the code combination is odd, or zero, if even. When decoding, the number of units in the code combination is calculated. If it turns out to be even, the received information is considered correct, if not, then erroneous.
Another form of error verification is the counting of checksum. This is a simple way that is commonly used with error controls using echoplex or parity / odd check. The essence of it is that the transmitting PEVM summarizes the numerical values \u200b\u200bof all transmitted characters. Sixteen youngest discharges of the amount are placed in a sixteen-digit counter of the checksum, which, together with the information of users, is transmitted by the receiving PC. The receiving PEVM performs the same calculations and compares the resulting checksum with transmitted. If these amounts coincide, it is understood that the block is transmitted without errors. The last word in the area of \u200b\u200bcontrol of errors in the field of PCM is a cyclic check with redundant code (CRC - Cyclic Redunduncy Check). It is widely used in HDLC protocols, SDLC, but in the PEVM industry appeared relatively recently. The error control field is included in the frame by the transmitting node. Its value is obtained as some function from the contents of all other fields. In the receiving node, identical calculations of another error control field are performed. These fields are then compared; If they match, the likelihood is that the package was transferred without errors.

Distribution of resources in networks
Web resources are very rich and continue to continuously replenish. This is web pages (containing text, images, java applets, frames, etc.), music files in MPZ format, recorded streaming audio and video, virtual worlds. Resources are distributed among a huge number of servers scattered around the world, and are available to millions of users. The HTTP protocol is a means that allows any user to get any object regardless of how many thousand kilometers is measured by the distance between the user's host and the remote server and how many Internet providers are on the request path. Nevertheless, the access time to Web resources is sometimes very significant. On the object path, the user's host has low-speed communication lines, which leads to significant transmission delays. On the path of the object there is at least one overloaded node, in which the value of the wait delay is great and there is a loss of packets. Overloads can occur even in cases where the node inputs are high-speed communication lines. The web server to which the query is addressed is overloaded, and the waiting time of the query service can be quite significant.
To solve the delay problem, a simple reception is used: the same resource is located on multiple servers, and the request is redirected to the "best" server. For a web page or MPZ file "The best" will be the server, the query time is minimal. Often, such a server belongs to the most close to the user host Internet provider.
The distribution of resources involves resource duplication mechanisms, as well as ways to determine the hosts of the servers that are most suitable for queries. In the second half of the 1990s, resource distribution tools were widespread; Currently, they actively apply, especially in the field of audio and video information. There are several large companies involved in the allocation of resources. Cisco, Lucent, Inktomi and Cacheflow are developing appropriate hardware and software, A AKAMAI, Digital Island and AT & T implement resource distribution services to resource providers such as Yahoo! and CNN. The distribution of resources is a field for active studies with both scientific and industrial points of view.
Over the years, engineers and researchers have proposed many decisions regarding the allocation of resources. These solutions can be approximately divided into three groups: Web caching, resource distribution network (CONTRITION NETWORKS, CDN) and peer-to-peer file separation. Below we will look at each of the technologies, however, I first clarify the terminology. We will assume any person, organization or company that have a resource available for Internet users. Under the source server of the object, the server will be implied on which the object was originally and where you can always find a copy of this object.
Web cache, often called a proxy server, is a network that performs http requests from the source server name. Web cache has its own information storage disk device containing previously requested copies of objects. As shown in Fig. The user browser can be configured in such a way that all the HTTP-requests created first are headed to the Web cache (this procedure in Microsoft and Netscape browsers is very simple).

After the browser is configured in the manner, any requested object is first searched for a Web cache. Typically, cache servers are rented and installed with Internet providers. For example, a university can create a cache server on its local network and configure all browsers so that they apply to the cache server.
Web caching is a resource allocation form, since duplicates source objects and organizes user access to local copies of objects. Note that the resource provider does not affect the duplication process; On the contrary, duplication depends only on user requests.
Caching was widely distributed on the Internet for three reasons. The first is that the cache servers are able to significantly reduce the execution time of the user's request, especially if the transfer rate between the user and the cache server exceeds the transmission rate between the user and the source server. Often, high-speed communication lines are used to connect the user with the cache server, so if there is a desired object on the cache server, its delivery to the user occurs in a very short time. The second reason for the popularity of caching mechanism is that it is capable of significantly reduced traffic between local networks and the Internet. This allows, in turn, reduce costs for expensive communication lines connecting local networks with the Internet. In addition, a significant reduction in traffic during caching occurs on the Internet as a whole, leading to better quality Services of applications of all users of the global network. Finally, the third cause of the success of caching is that it allows you to distribute resources among users at high speed. Even if the supplier uses low-speed low-speed network equipment, the most popular resources will soon be in Web caches, and, therefore, users will be able to download them with acceptable quality of service. Thus, the application of the cache server gives the best results than increasing the access bandwidth, and does not require replacement of network equipment. Of course, rental and installing the cache server is not free, however, the cost of the university in the event of a replacement of the access line would be significantly higher. Note that to create a Web cache is a fairly low-cost personal computer and, in addition, there are free software for cache servers.
Delivery Network (and Distribution) content (English Content Delivery Network or Content Distribution Network, CDN) is a geographically distributed network infrastructure that allows you to optimize the delivery and distribution of content to end users on the Internet. The use of CDN content providers contributes to an increase in the download speed of Internet users of audio, video, software, game, and other types of digital content at the point of the presence of the CDN network.
Delivery and distribution networks of content consist of geographically distributed multifunctional platforms, the interaction of which allows you to most effectively process and satisfy user requests when receiving content.
When using the CDN network, the data of the central Internet resource server is replicated to the peripheral platforms. Each platform supports a full or partial copy of the data distributed. The network node, which is part of the platform, interacts with local networks of Internet providers and distributes content to end users over the shortest network route with optimal server load. The length of the network route depends on the geographical or topological remoteness of the user computer from the server or the cost of transmitting traffic in the presence region.
Caching is the most common method of implementing the CDN solution, as it assumes the optimal use of disk space and connecting network channels. At the same time, the maximum cost of downloading the file (file queue) takes on the first user who turned to the original Content provider server. All subsequent users will access already downloaded replicas (HTTP objects) with the nearest server to them. Thus, only a popular and frequently requested content is stored on remote servers.
Large CDNs can consist of huge number distributed nodes and place their servers directly on the network of each local Internet provider. Many CDN operators focuse on the bandwidth of bonding channels and the minimum number of points of attachment in the presence region. Regardless of the architecture used, the main purpose of such networks is to accelerate the transmission of both static content and continuous data flow.
Depending on how the functions between network computers, network operating systems are distributed, and therefore, networks are divided into two classes: peer and double-aligned. If the computer provides its resources to other network users, then it plays the role of the server. At the same time, the computer appeals to the resources of another machine is a client. As already mentioned, a computer that works on the network can perform the functions of either the client or server, or combine both of these functions.
In peer-to-peer networks, all computers are equal to access to each other's resources. Each user may at wish to announce any resource of its computer shared, after which other users can exploit it. In such networks, all computers are installed on all computers, which provides all computers in the network potentially equal opportunities.
In peer-to-peer networks, a functional asymmetry may also occur: some users do not want to share their resources with others, and in this case their computers execute the client role, after other computers, the administrator consolidated only the functions for organizing the sharing of resources, which means they are servers, in the third The case when the local user does not object to the use of its resources and does not exclude the possibility of accessing other computers, the OS, installed on its computer, should include both server and client parts. Unlike networks with selected servers, in peer-to-peer networks there is no specialization of the OS depending on the prevailing functional direction - the client or server. All variations are implemented by means of configuring the same OS option.
Peer-to-peer networks are easier in the organization and operation, however, they are used mainly to combine small groups of users who do not have greater requirements for stored information, its protected from unauthorized access and access speeds. With enhanced requirements for these characteristics, double-to-one are suitable networks, where the server is better solving the task of servicing users with its resources, since its equipment and network operating system are specially designed for this purpose.

Protection and emergency recovery of information in the COP
Depending on the possible types of disorders of the network (under a violation of work, we also understand the unauthorized access) Numerous types of information protection are combined into two main classes:
- Physical protection tools, including means of protecting the cable system, power systems, archiving, disk arrays, etc.
- software protection, including: antivirus programs, deletion systems of authority, access control systems.
- Administrative protection measures, including room access control, development strategy of firm, emergency plans, etc.
It should be noted that such a division is quite conditionally, since modern technologies are developing towards a combination of software and hardware protection. Such software and hardware are most common, in particular, in the area of \u200b\u200baccess control, protection against viruses, etc.

Physical data protection

Cable system

The cable system remains the main "Achulese fifth" of most local computing networks: according to various studies, it is the cable system that causes more than half of all network failures. In connection with this cable system, special attention should be paid from the very moment of network design.
The best way to get rid of ourselves from the "headache" about the incorrect cable laying is the use of widely distributed so-called structured cable systems that use the same data cables in the local computer network, a local telephone network, video transmission or signals from the fire sensors. Safety or security systems. Structured cable systems include, for example, SYSTIMAX SCS AT & T Company,
The best way to protect the cable from physical (and sometimes temperature and chemical impacts, for example, in production workshops) is the cable laying using in varying degrees of protected boxes.
Another important problem of the proper installation and trouble-free operation of the cable system is the compliance of all its components with the requirements of international standards.

Power supply systems

The most reliable means of preventing information loss with short-term disconnection of electricity is currently the installation of sources. uninterrupted power. Different technical and consumer characteristics, such devices can provide power to the entire local network or a separate computer for a period of time sufficient to restore the voltage or to maintain information on magnetic media. Most uninterruptible power sources simultaneously perform functions and voltage stabilizers, which is an additional protection against voltage jumps in the network. Many modern network devices - servers, hubs, bridges, etc. - are equipped with its own dubbed power systems.
Abroad, large corporations have their own emergency electrical generators or backup power lines. These lines are connected to different substations, and at the failure of one of them, the power supply is carried out from the reserve substation.

Archiving and Duplication Information Systems

The organization of a reliable and efficient data archiving system is one of the most important tasks to ensure the safety of information on the network. In small networks where one or two servers are installed, the archiving system is most often used directly to the free slots of the servers. In large corporate networks, it is most preferable to organize a dedicated specialized archiving server.
Such a server automatically automatically archive information from hard drives of servers and workstations to the time-specified local computing network time, issuing a backup report. This ensures the management of the entire archiving process from the administrator console, for example, you can specify specific volumes, catalogs or separate filesthat need to archive. It is also possible to organize automatic persons on the occurrence of one or another event ("Event Driven Backup"), for example, when receiving information that there is little free space on the hard disk of the server or workstation, or when one of the "mirror "Discs on the file server. Among the most common archiving servers models, you can select Storage Express System Intel Corporation, ArcServe for windows, produced by Cheyenne and a number of others.
Storage of archival information representing a special value must be organized in a special protected room. Experts recommend to store duplicates of the archives of the most valuable data in another building, in case of a fire or natural disaster.
Improving the reliability and protection of data on a network based on the use of redundant information is implemented not only at the level of individual network elements, such as disk arrays, but at the level of the network OS.

Natural Disaster Protection

The main and most common method of protecting information and equipment from various natural disasters - fires, earthquakes, floods, etc. - is to store archive copies of information or place some network devices, such as database servers, in special protected premises located, As a rule, in other buildings or, less often, even in another area of \u200b\u200bthe city or other city.

Software and software and hardware protection methods

Protection against computer viruses

There are hardly at least one user or network administrator, which would never come across computer viruses. According to a study conducted by Creative Strategies Research, 64% of the 451 surveyed specialist experienced "on themselves" action of viruses. Today, in addition to thousands of already known viruses, 100-150 new strains appear monthly. The most common methods of protection against viruses to this day are various antivirus programs.
However, as a promising approach to protection against computer viruses in recent years, a combination of software and hardware protection methods is increasingly used. Among the hardware devices of such a plan, special anti-virus cards can be noted, which are inserted into standard computer expansion slots. Intel in 1994 has proposed a promising technology for protection against viruses in computer networks. Flash memory network adapters Intel EtherExpress Pro / 10 contains an antivirus program scanning all computer systems before loading it.

Safety from unauthorized access

The problem of protecting information from unauthorized access was particularly aggravated with the wide distribution of local and, especially global computer networks. It should also be noted that it is often damaged not because of the "malicious intent", but because of the elementary user errors who accidentally spoil or delete vital data. In this regard, in addition to controlling access, the necessary element of information protection in computer networks is the delimitation of user authority.
In computer networks, in the organization of access control and delimitation of user authority, built-in network operating systems are most commonly used.
One of the successful examples of creating a comprehensive access control solution open systemsBased on both software and hardware protection, the Kerberos system has become. The basis of this authorization scheme is three components:
- Database containing information on all network resources, users, passwords, encryption keys, etc.
- Authorization server (Authentication Server) processing all user requests for obtaining a particular type of network services. Authorization server, receiving a request from the user, refers to the database and determines whether the user has the right to perform this operation. It is noteworthy that user passwords over the network are not transmitted, which also increases the degree of information protection.
- Ticket-Granting Server (Resolution Server) receives from an authorization server "Skip", containing the username and its network address, request time and a number of other parameters, as well as a unique session key. A package containing "skip" is also transmitted in the deciduous DES encrypted algorithm. After receiving and decrypting "skip", the server issuing server checks the query and compares the keys and then gives "good" to the use of network equipment or programs.
Among other similar comprehensive schemes, it is possible to note the SESAME system developed by the European Association of Computers (ECMA). (Secure European System for Applications in Multivendor Environment) intended for use in large heterogeneous networks.

The local computing network (LAN) typically includes a set of user workstations, a network administrator workstation (one of the user stations can be used), the server kernel (a set of hardware server platforms with server programs: file server, www server, database server , mail server, etc.), communication equipment (routers, commutatrators, hubs) and structured cable system (cable equipment).

The reliability calculation of the LAN is starting with the formation of the concept of failure of this network. For this, management functions are analyzed, the execution of which in the enterprise is carried out using this LAN. These features are selected, the violation of which is unacceptable, and the LAN equipment is defined, involved in their execution. For example: Certainly during the working day, the ability to call / write information from the database, as well as access to the Internet.

For a combination of such functions according to the structural electrical circuit, the LAN equipment is determined, the refusal of which directly disrupts at least one of the specified functions, and a logical scheme for calculating reliability is drawn up.

At the same time, the quantities and conditions for the operation of repair and recovery brigades are taken into account. The following conditions are usually accepted:

Restoration limited - i.e. At any time, it cannot be recovered more than one failed element, because There is one repair brigade;

The average recovery time of the failed element is established or based on the permissible breaks in the work of the LAN, or from the technical capabilities of delivery and inclusion in the operation of this element.

Within the framework of the above approach to the calculation of the scheme for calculating reliability, as a rule, it can be reduced to a sequential-parallel scheme.

We establish a refusal of equipment included in the kernel of the network: servers, switches or cable equipment as a criterion to refuse LAN.

We believe that the failure of the workstations of users does not lead to the failure of the LAN, and since the simultaneous failure of all workstations is an unlikely event, the network during individual failures of workstations continues to function.

We will submit that the local network in question includes two servers (one provides Internet access), two switches and five cable fragments belonging to the network core. The intensity of failures and recovery for them is shown below, are still k \u003d 1-L / m.

The values \u200b\u200bof the intensity of recovery are maximal for cables, the replacement of which is carried out using the spare and minimal for switches, the repair of which is carried out by specialized firms.

Calculation of the characteristics of the subsystems of servers, switches and cables is carried out by expressions for the serial connection of the elements.

Server subsystem:

l C \u003d 2 * L 1 \u003d 4 * 10 -5; To GS \u003d 1-4 * 10 -4; M C \u003d 1 / h.

Switch subsystem:

l k \u003d 2 * 10 -5; To GK \u003d 1-2 * 10 -3; M K \u003d 1 / h.

Cable subsystem:

l l \u003d 5 * 10 -6; To ch \u003d 1-5 * 10 -6; M l \u003d 1 / h.

For the entire network:

l S \u003d 6.5 * 10 -5; K g s \u003d 1-2.4 * 10 -3; M s \u003d 0.027 1 / h.

Calculation result:

T \u003d 15 thousand h., K g \u003d 0.998, T B "37 h.