Protocol for connecting computers on a local network. Protocols and technologies of local networks. Improving the performance of local networks

Interaction between the same levels of the model in different subscriber computers must be carried out according to certain rules.

Protocol is a set of rules that determines the interaction of two levels of the same name in the open systems interaction model (OSI) in various subscriber computers.

A protocol is not a program. The rules and sequence of actions during information exchange, defined by the protocol, must be implemented in the program.

The protocols of the three lower levels of the open systems architecture model are the easiest to standardize, since they define the actions and procedures characteristic of computer networks of any class.

The most difficult thing to standardize is the protocols of the upper levels, especially the application one, due to the multiplicity applied problems and in some cases their uniqueness. If, based on the types of structures, methods of access to the physical transmission medium, the network technologies used and some other features, one can count about a dozen various models computer networks, there are no limits to their functional purpose.

Internet. Construction principles

Internet- connection, integration of various networks. When combining several networks into a single whole, a special internetworking protocol is used. Internet Protocol, in English - Internet Protocol (IP) , and gave the name to the Internet. Data exchange in this global network is carried out using the Data Transmission Control Protocol - Transmission Control Protocol (TCP) . Thus, information is transported to the Internet using the so-called TCP/IP protocol stack.

Internet- a global computer network connecting individual networks that operate according to the TCP/IP protocol (Transmission Control Protocol / Internet Protocol), are united through gateways and use a single address space and name space.

At the heart of the Internet is a system of backbone networks called backbones. Regional mid-level networks provide connection of individual territories to a high-speed core network.

On the Internet, data exchange between nodes can be carried out along different routes, along different communication lines. The failure of an individual telecommunications channel does not lead to a complete loss of communication.

History of the Internet

In the 60s of the twentieth century, after the Cuban missile crisis, specialists rand Corporation (US think tank) proposed creating a decentralized computer network covering the entire country. The idea was that even in the event of a nuclear attack, the connection between military computers of scientific and educational institutions connected to this network would not be destroyed. Such a structure could only be implemented if multiple connections exist between network nodes, that is, all nodes must have the same status, each node is authorized to generate, transmit and receive messages from any other node. If in a normal network a server, when it failed, brought down the entire network, then in new network, there should have been an arbitrary number of servers, each of which could choose the path for sending information.

It was assumed that data intended for transmission should be broken down into small standard blocks of data called packages. Each packet must have a destination address and its delivery is ensured by the fact that each node has the ability to send (or forward) packets across the network to its destination.

In 1968, one of the divisions of the Pentagon, the ARPA agency, began funding this project and in the fall of 1969, the ARPANET network appeared, consisting of only four nodes: SDS SIGMA (University of California), SDS-940 (Stanford Research Institute), IBM-360 ( California Institute of Santa Barbara), DEC PDP-10 (University of Utah). The birthday of the Internet is considered to be October 29, 1969, when the first attempt was made to remotely connect to a computer at the Stanford University research center from another computer at the University of California at Los Angeles. Vinton Cerf is sometimes called the "Founding Father" of the Internet. In 1971, there were already 15 nodes, and in 1972 - 37 nodes. In 1973, foreign nodes were connected to the network - in London and Norway. In 1974, the NSF (National Science Foundation) published the TCP/IP protocol standard, which became standard on the ARPANET in 1983. By this year, the network already had the established name INTERNET. The Internet began to grow rapidly in the 1980s. The scheme of connecting computers into a network with decentralized control has spread throughout the world.

A network protocol is a set of rules that allows connection and data exchange between two or more computers connected to a network. In fact, different protocols often describe only different aspects of the same type of communication; taken together, they form the so-called protocol stack. Titles<протокол>And<стек протоколов>also indicate the software that implements the protocol

Protocol levels

The most common classification system for network protocols is the so-called OSI model. In accordance with it, protocols are divided into 7 levels according to their purpose - from physical (generation and recognition of electrical or other signals) to application (API for transferring information by applications):

• Application layer. The upper (7th) level of the model ensures interaction between the network and the user. The layer allows user applications to access network services such as database query handler, file access, forwarding Email. It is also responsible for transmitting service information, providing applications with information about errors, and generating requests to the presentation layer. Example: HTTP, POP3, SMTP.
• Presentation layer. Layer 6 is responsible for protocol conversion and data encoding/decoding. It converts application requests received from the application layer into a format for transmission over the network, and converts data received from the network into a format that applications can understand. The presentation layer can perform compression/decompression or encoding/decoding of data, as well as redirecting requests to another network resource if they cannot be processed locally.
• Session layer. Level 5 of the model is responsible for maintaining a communication session, which allows applications to interact with each other for a long time. The session layer manages session creation/termination, information exchange, task synchronization, determination of data transfer rights, and session maintenance during periods of application inactivity. Transmission synchronization is ensured by placing it in the data stream control points, starting from which the process is resumed if interaction is disrupted.
• Transport layer. The 4th level of the model is designed to deliver data without errors, losses and duplication in the sequence in which they were transmitted. It does not matter what data is transmitted, from where and where, that is, it provides the transmission mechanism itself. It divides data blocks into fragments, the size of which depends on the protocol, combines short ones into one, and splits long ones. Protocols at this level are designed for point-to-point communication. Example: TCP, UDP
• Network layer. Layer 3 of the OSI network model is designed to determine the data transmission path. Responsible for translating logical addresses and names into physical ones, determining the shortest routes, switching and routing, monitoring problems and congestion in the network. A network device such as a router operates at this level.
• Data Link layer. This level is often called the channel level. This layer is designed to ensure the interaction of networks at the physical layer and control errors that may occur. It packs the data received from the physical layer into frames, checks for integrity, corrects errors if necessary, and sends it to the network layer. The data link layer can communicate with one or more physical layers, monitoring and managing this interaction. The IEEE 802 specification divides this layer into 2 sublayers - MAC (Media Access Control) regulates access to the shared physical medium, LLC (Logical Link Control) provides network layer service. Switches and bridges operate at this level. In programming, this level represents the driver network card, V operating systems there is a software interface for the interaction of the channel and network layers with each other, this is not new level, but simply an implementation of the model for a specific OS. Examples of such interfaces: ODI, NDIS
• Physical layer. The lowest level of the model is intended directly for transmitting the data stream. Transmits electrical or optical signals into a cable or radio broadcast and, accordingly, receives and converts them into data bits in accordance with digital signal encoding methods. In other words, it provides an interface between the network media and the network device. At this level, signal concentrators (hubs), signal repeaters (repeaters) and media converters operate. Physical layer functions are implemented on all devices connected to the network. On the computer side, the physical layer functions are performed by the network adapter or serial port.

The TCP/IP protocol is mainly used

Computer Network Terminology Protocols

A network protocol is a set of software-implemented rules for communication between computers connected to a network. It is practically the “language” in which computers talk to each other. Currently, the standard has become the use of only the TCP/IP protocol. In previous Windows versions Several protocols were installed by default, usually NetBEUI, NWLink IPX/SPX, TCP/IP.

• NetBEUI.
  Compact and efficient protocol for interaction in small networks (up to 200 computers). Used in a wide variety of systems: Microsoft LAN Manager, Windows 3.1/3.11 for Workgroups/95/98/NT 4.0, IBM PCLAN, LAN Server, etc. In Windows 2000 and older, a new specification of this protocol is used, which is called NetBIOS Frame Protocol (NBFP). NetBEUI (NBFP) does not require any additional settings. If you need to quickly create a network and you do not feel confident in understanding the additional settings that, for example, the TCP/IP protocol requires, then enable the NBFP protocol. You will get a simple and very fast functioning local network.
• NWLink IPX/SPX.
  If there are Novell NetWare servers on the network, then this protocol is necessary to organize communication with them. Otherwise, this protocol should be excluded from those used in the system.
• TCP/IP.
  The main recommended protocol for both large enterprise networks and small offices, and for connecting home computers into a private network. Unlike other protocols, it requires a number of preliminary settings.

Note
You should not use more services and protocols on your network than are required for normal operation in a specific situation. Firstly, this will unproductively use computer resources. Secondly, any additional service and unused protocol is another “entrance” to the system that must be protected. Therefore it is easier not to provide additional features hackers than to constantly monitor vulnerabilities found in these services, install the necessary updates, etc.

OSI model

For the purpose of systematization, the OSI model is often used, which conventionally divides network interaction into seven layers.
Knowledge of OSI levels is usually required when passing certain certification exams, but in practice this division has lost its meaning. If the first three levels can still be isolated quite well when analyzing this or that network project, then it is quite difficult to classify the functionality of the equipment according to other levels. For marketing purposes, it is often indicated in the descriptions of switches that they operate, for example, at level 4 or 7. In practice, this only means that when implementing certain functionality in switches, the data packet is analyzed according to characteristics related to the corresponding levels. For example, this happens during multicast traffic routing operations (the switch analyzes the packet to determine whether it belongs to a particular program), packet prioritization, etc.

TCP/IP protocol stack

When they talk about TCP/IP, then this name usually means many different protocols based on TCP/IP. There are a large number of different standards that define certain options for interaction on the Internet.
So, there are rules by which messages are exchanged between mail servers, and there are rules by which end users can receive letters in their mailbox. There are rules for conducting video and audio broadcasts, and rules for organizing telephone conversations over the Internet. There are rules that determine the behavior of data transfer participants in the event of an error, etc.
It is logical that when developing file transfer rules, no one creates new mechanisms for transferring a single data packet and that the file transfer protocol is based on a simpler packet transfer protocol.

Therefore, it is customary to say that there are layers of the IP protocol, and at each layer there are various variants of special protocols. This entire set of protocols is called the TCP/IP protocol stack.

Protocols UPD, TCP, ICMP

Protocols are used to transmit data TCP(Transmission Control Protocol) and UDP (User Data gram Protocol, user datagram protocol). UDP used in cases where reception confirmation is not required (for example, DNS queries, IP telephony). Data transfer via protocol TCP provides for confirmation of receipt of information. If the transmitting party does not receive the required confirmation within the specified time frame, the data will be retransmitted. Therefore the protocol TCP are referred to as connection oriented protocols, a UDP- no (connection less).
Internet Control Message Protocol ( ICMP, Internet Control Message Protocol) is used to transmit network parameter data. It includes packet types such as ping, destination unreachable, TTL exceeded, etc.

The rapid development of the Internet has led to the fact that the parameters laid down when creating IP protocols began to restrain the further development of the global network. Therefore, numerous groups are constantly developing possible modifications to this protocol. The most “recognized” development at the moment is considered to be a project of the IETF group (Internet Engineering Task Force, a problem group for Internet design), which is called IPv6 (other projects are collectively called IP Next Generation or IPng).
To the main features this project relate:

• keeping the basic operating principles of the IP protocol unchanged;
• use of longer addresses (128-bit);
• use of built-in 64-bit encryption algorithm;
• taking into account the mechanism for reserving protocol bandwidth (previously the problem was solved by introducing classes of service);
• the presence of great opportunities for further expansion of functions: only part of the characteristics is strictly described, the rest allow for further development.

Although the majority of Internet participants support the development of this protocol, the actual implementation of this development will require a long time and significant investment, since it entails upgrading a large amount of already installed equipment.
Support for the IPv6 protocol is built into Windows operating systems, starting with Windows XP. To enable it in Windows XP, you must run the ipv6 install command. But the use of ipv6 is not yet practical. According to various estimates, a shortage of IPv4 protocol address space may occur no earlier than in 5-10 years. This is sufficient time for the development of the next IP protocol specification.

TCP/IP protocol parameters
IP address

Each computer operating using the TCP/IP protocol must have an IP address - a 32-bit number used to identify a node (computer) on the network. It is customary to write the address in decimal values ​​of each octet of this number, separating the resulting values ​​with dots. For example: 192.168.101.36.
IP addresses are unique. This means that each computer has its own combination of numbers, and there cannot be two computers on the network with the same addresses. IP addresses are distributed centrally. Internet providers make applications to national centers in accordance with their needs. The address ranges received by providers are further distributed among clients. Clients themselves can act as an Internet provider and distribute the received IP addresses between subclients, etc. With this method of IP address distribution computer system knows exactly the “location” of a computer that has a unique IP address; All she needs to do is send the data to the “owner’s” network. The provider, in turn, will analyze the destination and, knowing who is given this part of the addresses, will send the information to the next owner of the IP address subrange until the data reaches the destination computer.
Allocation of a range of addresses is free, but the organization receiving the addresses must actually confirm their use after a certain period of time.
To build local networks of organizations, special ranges of address zones are allocated. These are the addresses Yu.x.x.x, 192.168.x.x, Yu.x.x.x, from 172.16.x.x to 172.31.x.x, 169.254.x.x. Packets sent from the specified addresses are not routed (in other words, not forwarded) over the Internet, so computers on different local networks may have matching addresses from the specified ranges. To send information from such computers to the Internet and back, special programs are used that “on the fly” replace local addresses with real ones when working with the Internet. In other words, data is sent to the Network from a real IP address. This process occurs "invisibly" to the user. This technology is called address translation.

Group addresses

If data must be transferred to several devices (for example, viewing video from one Web camera on different computers or simultaneously deploying an operating system image to several systems), then using group broadcasts can reduce the load on the network.
To do this, the computer is assigned another IP address from a special range: from 224.0.0.0 to 239.255.255.255, and the ranges 224.0.0.0-224.0.0.255 and 239.0.0.0-239.255.255.255 cannot be used in applications and are intended for routing protocols3 etc. Group mailing addresses are assigned using the appropriate software.
If the switch has functions for working with multicasts (support for IGMP snooping, P1M DM/PIM SM), then data transmitted to multicast addresses will only arrive on those ports to which devices that have subscribed to the corresponding broadcasts are connected. As a result, network traffic can be significantly reduced compared to the option of transmitting such data to each network device independently.

Distribution of IP addresses for a small office network

Enterprise networks typically use IP address ranges that are allocated for local use. Some addresses are assigned statically, others are distributed dynamically using DHCP (Dynamic Host Configuration Protocol, dynamic server configuration protocol).

Static addresses are assigned:

• behind the gateway, for which the address xxx.xxx.xxx.1 is usually used, but this is a tradition, not a rule;
• behind DNS servers, DHCP, WINS;
• behind domain controllers;
• behind network servers (for example, centralized file shares, mail server and so on.);
• at printing stations that have a direct connection to the network;
• behind managed network devices (for example, network switches, SNMP-managed emergency power supplies, etc.).

Workstations traditionally use dynamic addresses. Moreover, part dynamic addresses is issued for local use, and part is intended for external clients, “guests” of the network.

Note
Typically, computers that receive guest addresses are subject to certain restrictions on access rights to internal resources.

Address mask

The concept of a subnet was introduced so that it is possible to select part of the IP addresses of one organization, part of another, etc. A subnet is a range of IP addresses that are considered to belong to one local network. When working on a local network, information is sent directly to the recipient. If the data is intended for a computer with an IP address that does not belong to the local network, then special rules are applied to it to calculate the forwarding route from one network to another. Therefore, when using the TCP/IP protocol, it is important to know which network the recipient of the information belongs to: local or remote.
A mask is a parameter that “tells” the software how many computers are included in a given group (“subnet”). The address mask has the same structure as the IP address itself: it is a set of four groups of numbers, each of which can be in the range from 0 to 255. Moreover, the lower the mask value, the more computers united into this subnet. For small business networks, the mask is usually 255.255.255.x (for example, 255.255.255.224). The network mask is assigned to the computer along with the IP address.

So, network 192.168.0.0 with a mask of 255.255.255.0 (otherwise you can write 192.168.0.0/24) can contain hosts with addresses from 192.168.0.1 to 192.168.0.254. The address 192.168.0.255 is the broadcast address for this network. And the network 192.168.0.0 with a mask of 255.255.255.128 (192.168.0.0/25) allows addresses from 192.168.0.1 to 192.168.0.127 (the address 192.168.0.128 is used as a broadcast address).
In practice, networks with a small possible number of hosts are used by Internet providers (in order to save IP addresses). For example, a client might be assigned an address with a mask of 255.255.255.252. This subnet contains only two hosts. When partitioning a network, organizations use local address ranges for Class C networks. A Class C network has an address mask of 255.255.255.0 and can contain up to 254 hosts. The use of class C networks when divided into VLANs in an enterprise environment is due to the fact that automatic routing protocols use precisely such subnets.
When creating subnets in an organization, it is recommended to adhere to the following rule: subnets related to a specific distribution node must be part of the same network. This simplifies routing tables and saves switch resources. For example, if the subnets 192.168.0.0/255.255.255.0, 192.168.1.0/255.255.255.0, 192.168.3.0/255.255.255.0 are connected to this switch, then it is enough for the other switch to know that packets for the 192.168 network should be forwarded in this direction .0.0 /255.255.252.0.
This recommendation is not significant for networks of small and medium-sized organizations, since the resources of modern switches are sufficient to store settings of such a volume.

After the computer has received an IP address and “knows” the value of the subnet mask, the program can begin working in this local subnet. To exchange information with other computers on the global network, you need to know the rules of where to send information to the external network. For this purpose, such a characteristic of the IP protocol as the gateway address is used.

Gateway (default gateway)

Gateway(gateway) is a device (computer) that provides information transfer between different IP subnets. If the program determines (by IP address and mask) that the destination address is not part of the local subnet, then it sends this data to the device that acts as a gateway. In the protocol settings, specify the IP address of such a device.
A gateway may not be assigned to operate only on a local network.
For individual users connecting to the Internet, or for small businesses with a single connection channel, the system should have only one gateway address - this is the address of the device that has a connection to the Network. If there are multiple routes (paths for sending data to other networks), there will be multiple gateways. In this case, a routing table is used to determine the data path.

Routing tables

An organization may have several points of connection to the Internet (for example, for the purpose of reserving data transmission channels or using cheaper channels, etc.) or contain several IP networks in its structure. In this case, so that the system “knows” which way (through which gateway) to send this or that information, routing tables are used. The routing tables for each gateway indicate those Internet subnets for which information should be transmitted through them. In this case, for several gateways you can set the same destination ranges, but with at different prices data transmission: information will be sent over the channel that has the lowest cost, and if it fails for one reason or another, the next “cheapest” connection will be automatically used.
Routing tables exist on every device that uses the IP protocol. Administrators primarily work with switching equipment routing tables. Setting up computer routing tables makes sense only if there are several network adapters connected to different network segments. If the computer has only one network card (one connection to the Internet), the routing table has the simplest form: it states that all signals should be sent to the default gateway.

You can view the TCP/IP protocol routing table using the route print command. Using the route command, you can also add a new static route (route add) or a permanent route - route add -p (the route is saved in the settings after a system reboot).
Let's show with an example how modifications to the routing table can be used. Let's assume that the computer has two network cards, one of which is directly connected to the Internet (has real address), and the second is used to work on the internal network (local address). Internet access is provided by default through a gateway on the local network. In this case, the routing table displayed by the route print command looks something like this:

Let's check the path of packets to an Internet address, for example 109.84.231.210, using the tracert command:
tracert 109.84.231.210 -d As a result, we get something like this (the listing is limited to the first four nodes):

Let's say we want to change the path of packets to a host of our choice by routing the information through a second network card (rather than through the default gateway). To do this, use the route add command to add the route we want:
route add 109.84.231.210 mask 255.255.255.255 195.161.192.2
In the command, we indicated that we wanted to assign a new route not for a range of addresses, but only for a specific value (therefore the mask is 255.255.255.255). In addition, they explicitly specified the address of the network interface through which packets should be forwarded.
After executing this command (the system does not display any operation results), the changes can be viewed through the routing table.

Compared with original version the routing table was supplemented with one line, which is shown in this example (the remaining lines have not changed).

Checking the new signal path:
Route tracing to 109.84.231.210 with a maximum number of hops of 30

1 1ms 1ms 1ms 195.161.192.1
2 23 ms 22 ms 23 ms 195.161.94.137
3 23 ms 23 ms 23 ms 195.161.94.5

...
It can be seen that the packets are sent through a different interface.
These routing changes last until the system is rebooted or until the reverse command is issued: deleting routing entries. To restore routing parameters, just issue a command specifying the route that you want to delete:

route delete 109.84.231.210

In this case, it is usually possible not to specify the mask and interface parameters (if they are uniquely determined by the address entered in the command).

Note
In practice, there are situations when changing routing parameters in the Windows operating system was not immediately processed correctly. Sometimes, after operations on the routing table, in order to achieve success, it was necessary to programmatically disable and re-enable the network interface for which the configuration was being performed.

Understanding routing rules is important not only when building routes on the Internet, a task that network administrators of small enterprises are unlikely to have to solve. In practice, virtual networks are widely used to isolate separate sections of a local network (for example, for security reasons). And in order to provide selective access to such networks, administrators must be able to write the correct routing table for the corresponding VLAN.

Chapter 5

Local network protocols

After reading this chapter and completing the practice exercises, you will be able to:

Ø Explain the following protocols and their use in various network operating systems:

Ø discuss and implement methods to improve the performance of local networks.

At the beginning of the 20th century, sociologist George Herbert Mead, studying the influence of language on people, came to the conclusion that human intelligence primarily developed through language. Language helps us find meaning in the surrounding reality and interpret its details. In networks, a similar role is played by network protocols, which allow diverse systems to find a common environment for interaction.

This chapter describes the protocols most commonly used on local area networks and the network operating systems that use them. You will learn about the advantages and disadvantages of each protocol, which will help you understand their uses. The most popular local network protocol, TCP/IP, is discussed only briefly in this chapter, since it will be described in more detail in Chapter 6. At the end of this chapter, you will be introduced to methods for improving the performance of local networks and selecting the protocols that are needed in a particular situation.

Local network protocolsand their application in networksoperating systems

Network protocols are like a local language or dialect: they enable networks to seamlessly exchange information between connected devices. These protocols are also important for simple electrical signals transmitted over a network communication cable. I protocol network communications would simply be impossible. In order for two computers to communicate freely with each other, they must use the same protocol, just as two people must communicate in the same language. I

In a local network, several protocols can operate individually and in some combinations. Network devices (such as routers) are often configured to automatically recognize and configure different protocols (depending on the operating system used in the router). For example, on a single Ethernet LAN, one protocol might be used to connect to the mainframe, another to work with Novell NetWare servers, and a third to work with Windows servers (for example, running Windows NT Server) (Figure 5.1).

You can install a bridge router that will automatically recognize each protocol and configure itself accordingly, causing it to act as a router for some protocols and as a bridge for others. The presence of several protocols in a network is effective in that such a network can simultaneously perform many functions (for example, provide Internet access to mainframes and servers). The disadvantage of this approach is that some protocols will operate in broadcast mode, that is, they will periodically send packets to identify network devices, generating significant excess traffic.

Some network protocols have become widely used because they are associated with specific network operating systems (for example, Windows systems, IBM mainframes, UNIX servers, and Novell NetWare). It makes sense to study protocols in relation to the operating systems where they are used. In this case, it becomes clear why a specific protocol is needed in a certain type of network. It will also make it easier for you to understand how one protocol (such as NetBEUI) can be replaced by other protocols (such as TCP/IP). However, before learning about protocols and their interrelationships between operating systems, it is important to learn about the general properties of LAN protocols.

General propertieslocal network protocols

Basically LAN protocols have the same properties as others communication protocols, however, some of them were developed long ago, during the creation of the first networks, which were slow, unreliable and more susceptible to electromagnetic and radio interference. Therefore, some protocols are not entirely suitable for modern communications. The disadvantages of such protocols include weak defense from errors or excessive network traffic. In addition, certain protocols were created for small local networks and long before the advent of modern corporate networks with advanced routing capabilities.

Local network protocols must have the following basic characteristics:

Ensure the reliability of network channels;

Have high performance;

Process source and destination node addresses;

Comply with networking standards, especially IEEE 802.

In general, all the protocols discussed in this chapter meet these requirements, but, as you will learn later, some protocols have more capabilities than others.

In table 5.1 lists local network protocols and operating systems with which these protocols can work. Later in the chapter, protocols and systems (in particular, server operating systems and host computers) will be described in more detail.

4 Table 5.1. Local network protocols and network operating systems

Protocol	Corresponding Operating System
	The first versions of Microsoft Windows operating systems
	UNIX, Novel NetWare, modern versions of Microsoft Windows operating systems, IBM mainframe operating systems
	IBM mainframe and minicomputer operating systems
	Client systems interacting with IBM mainframes configured to work with the SNA protocol

Note

Computer operating system is a set of software that performs two functions on a computer. First, they interact with the computer's hardware and the Basic input/output system (BIOS). Secondly, they interact with user interface(for example, with a graphical user interface (GUI) Windows systems or with the X Window Subsystem and desktops on UNIX systems). For network computer operating systems There is a third level of interaction in which these systems can communicate with each other over a network using one or more protocols.

ProtocolsIPX/ SPX and systemNovell NetWare

Protocol Internetwork Packet Exchange (IPX) (internetwork packet exchange) was developed by Novell for one of the very first network operating systems that performs server functions, called NetWare. This system was originally intended for Ethernet bus networks, token ring networks, and ARCnet networks, and was designed to work with a single file server. ARCnet is one of the proprietary alternative network technologies that uses special packages with markers and mixed topology (bus and star). Currently, the NetWare operating system has become hardware independent and can support various topologies and protocols.

As a prototype for the IPX protocol, Novell used one of the first local network protocols, the IPX protocol. Xerox Network System (XNS), adapting it for its file server operating system NetWare. Xerox Corporation proposed the XNS protocol as a means of transmitting data over Ethernet networks. In the early 1980s, several manufacturers released their own versions of this protocol. Novell's version spawned the IPX protocol for NetWare servers. At the same time, this company developed a companion protocol called Sequenced Packet Exchange (SPX) and focused on working with application programs, such as databases.

The IPX/SPX protocols are widely used in NetWare servers up to and including version 4. Beginning with NetWare 5.0, Novell is encouraging users to migrate to the TCP/IP protocol stack. These protocols are currently the primary protocols for NetWare 6.0 and later, although users may continue to use IPX/SPX protocols, particularly for compatibility with legacy servers and equipment (such as printers).

When IPX/SPX protocols are configured on an Ethernet network based on NetWare servers, four types of Ethernet frames can be used:

o 802 .2 - relatively new type frames used in networks based on NetWare servers versions from 3.21 to 4.x;

o 802.3 – an old frame type used on NetWare 286 systems (versions 2.x) and the first versions of the NetWare system and 3.1x);

o Ethernet II – to ensure compatibility with Ethernet II networks and more efficient frame formatting;

o Ethernet SNAP – implementation described in chapter 2 SubNetwork Access Protocol (SNAP), designed for the operation of special networks and applications from manufacturers.

Advantages and disadvantages

The advantage of the IPX protocol (despite its advanced age) compared to other early protocols is the possibility of its routing, i.e., the fact that it can be used to transmit data over many subnets within an enterprise. The disadvantage of the protocol is additional traffic, which occurs because active workstations use frequently generated broadcast packets to confirm their presence on the network. With many NetWare servers and hundreds of clients, IPX's "I'm here" broadcasts can generate significant network traffic (Figure 5.2).

Purpose of the SPX protocol

The SPX protocol, which complements IPX, provides data transmission application programs with greater reliability than IPX. IPX is slightly faster than its companion protocol, but it uses connectionless services running in the LLC sublayer of the Link Layer. This means that IPX guarantees that the frame will be delivered to its destination with a lower probability. The SPX protocol uses connection-oriented services, which improves the reliability of data transmission. Most often, when referring to both protocols (IPX and SPX), the abbreviation IPX/SPX is used.

The SPX protocol is widely used to transmit data content over the network. In addition, Novell's Remote Console Utility and Print Services operate based on this protocol. The remote console allows the administrator's workstation to see the same information that is displayed on the NetWare file server console, allowing the user to remotely execute system commands on the server without having to be at the server's keyboard.

Protocol DeploymentIPX/ SPX

To install the IPX/SPX protocols on computers running DOS, special DOS drivers developed for NetWare are used. On 32-bit operating systems (for example, Windows 95 and older), to install protocols, you can run the Novell Client32 program, which provides a command environment for accessing NetWare servers.

To enable computers running Windows systems to access NetWare, you can also use two types of drivers that allow you to work with several protocols: Open Datalink Interface (ODI) and Network Driver Interface Specification (NDIS).

When multiple protocols (such as IPX/SPX and TCP/IP) are deployed on a NetWare network, servers and clients often use a driver Open Datalink Interface, ODI(open channel interface). This driver enables communication with NetWare file servers, mainframes and minicomputers, as well as with the Internet. ODI drivers can be used in network clients running under MS-DOS and Microsoft Windows.

In earlier versions of Windows (Windows 3.11, Windows 95, Windows 98, and Windows NT), Microsoft implemented the GDI driver as a 16-bit application that could not take full advantage of the performance and capabilities of 32-bit Windows 95 and later.

Starting with Windows 95, more advanced solutions from Microsoft are used to connect to NetWare servers via the IPX/SPX protocol - protocol NetWare Link (NWLink) IPX/ SPX and driver Network Driver Interface Specification, NDIS(Network Adapter Standard Interface Specification). Practice Exercises 5-1 and 5-2 show you how to configure Windows 2000 and Windows XP Professional systems to use the NWLink protocol.

As shown in Fig. 5.3, NDIS (Microsoft) and ODI (Novell) drivers operate at the LLC sublayer of the Data Link layer, however, only one of these drivers can be bound to a network adapter at a time.

DIV_ADBLOCK20">

EmulationIPX/ SPX

The NWLink protocol emulates IPX/SPX operation, so any Windows system that uses it operates as a computer or device configured for IPX/SPX. NDIS is a driver software specification used by the NWLink protocol that allows it and other network protocols to communicate with a computer's network adapter. This uses a procedure to establish communication between the protocol and the adapter, called binding. Binding(binding) a certain protocol to a specific adapter allows this adapter to operate and provide an interface with the network environment.

Binding to the driverNDIS

The Microsoft NDIS driver can bind one or more protocols to a single network adapter, allowing all of those protocols to work through that adapter. If there are several protocols, then a certain hierarchy is established between them, and if several protocols are deployed on the network, then the network adapter will first try to read the frame or packet using the protocol located at the top level of this hierarchy. If the formatting of the frame or packet corresponds to a different protocol, then the adapter will try to read it using the next protocol specified in the hierarchy, and so on.

Advice

Using the NDIS driver, one protocol can be bound to several network adapters on a computer (for example, on a server). If you have several adapters, you can distribute the network load between them and speed up the server's response to requests when there are a large number of users. In addition, multiple adapters are used if the server also functions as a router. Binding one protocol to multiple adapters also reduces memory footprint because the server does not need to load multiple instances of the same protocol into it.

It should be noted that the user can independently organize the hierarchy of protocols associated with the adapter. This hierarchy is called the binding order. For example, if the first protocol in the hierarchy is IPX/SPX and the second is TCP/IP, then the TCP/IP frame or packet is first interpreted as IPX/SPX data. The network adapter quickly detects the error and rereads the TCP/IP frame or packet, recognizing it correctly.

The protocol binding order can be set in most Microsoft Windows operating systems (for example, Windows 2000 and Windows XP). In Fig. Figure 5.4 shows the binding procedure on a computer running Windows XP Professional. In this figure, the protocols are listed below the line File and Printer Sharing for Microsoft Networks, display nil doc bindings for protocols used to access shared files and printers. Below the line Client for Microsoft Networks shows the order of binding protocols required to access network servers. In Practice Exercises 5-3 and 5-4, you will learn how to set the protocol binding order in Windows 2000 and Windows XP Professional.

DIV_ADBLOCK22">

Note

As discussed earlier in this book, it is not recommended to enable RIP on NetWare and Windows 2000/Server 2003 servers because it introduces additional traffic on the network. It is preferable for specialized network routers to perform all routing tasks.

Table 5.2. Protocols used with serversNetWare

Abbreviatura	Full title	Description	LevelmodelsOSI
	Internetwork Packet Exchange	Used as the primary data transfer protocol for Ethernet applications. All frame types can be used: Ethernet 802.2, Ethernet 802.3, Ethernet II and Ethernet SNAP	Network and Transport
	Link Support Layer	Used in conjunction with the ODI driver to support multiple protocols on a single network adapter	Duct
	Multiple Link Interface Driver	Connects two or more channels into one telecommunications line (for example, two ISDN terminal adapters). In Ethernet networks, the MLID protocol in combination with the workstation network adapter allows you to determine the level of conflicts in the network; in networks with a token ring, it coordinates token transfers	Channel (MAC sublayer)
	NetWare Core Protocol	Part of the operating system that facilitates communication between clients and servers when accessing applications or open files located on a NetWare server
	NetWare Link Services Protocol	Provides IPX packets with routing information
	Routing Information Protocol	Collects routing information for servers that provide routing services
	Service Advertising Protocol	Allows NetWare clients to identify the servers and network services running on them. Servers generate SAP broadcast packets every 60 s, and clients use them to locate the nearest server	Session Executive Application
	Sequenced Packet Exchange	Provides application programs with a connection-oriented data transfer mechanism	Transport

ProtocolNetBEUI and serversMicrosoft Windows

Microsoft Windows NT began as a joint project between Microsoft and IBM to develop the LAN Manager server operating system. In the early 1990s, Microsoft transitioned from LAN Manager to Windows NT Server, which later became a widely used operating system.

Based on the Windows NT Server product, Windows 2000 Server and Windows Server 2003 were created. Like modern versions of Novell NetWare, Windows NT, Windows 2000 and Windows Server 2003 systems are compatible with Ethernet and Token Ring local networks, they can scale from small computers from Intel-compatible processors to multiprocessor systems. Most often, TCP/IP protocols are used with these systems, but Windows NT systems are still available Server versions 3.51 and 4.0, which implement the native protocol of Windows NT systems - NetBIOS Extended User Interface, NetBEUI. This protocol was created for the LAN Manager and LAN Server operating systems before Windows came into existence BEUI was implemented in the first versions of Windows NT still available in Windows 2000 (although no longer supported in Microsoft systems, starting with Windows XP).

Note

On computers running Windows NT and Windows 2000, the NetBEUI protocol is also found under the name NBF (NetBEUI frame). If you use a protocol analyzer to analyze network traffic, then NetBEUI frames will be marked with exactly this abbreviation.

StoryNetBEUI

The NetBEUI protocol was originally developed by IBM in 1985 as an improved modification Network Basic Input/ Output System, NetBIOS(basic network input/output system). NetBIOS is not a protocol, but a method for application programs to interact with network devices, as well as a name recognition service used on networks. BIOS names are given to various network objects (such as workstations, servers, or printers). For example, a username can be used to identify his workstation on a network, HPLaser can be used to access a network printer, and a server can be named AccountServer. Such names make it easier to find the necessary network resources. They are translated (converted) into addresses used in network communications using NetBIOS Name Query services.

Application areaNetBEUI

The NetBEUI protocol was developed at a time when computer networks primarily meant local area networks for a relatively small number of computers (from a few to two hundred). The design process did not take into account the features of corporate networks with packet routing. For this reason, the NetBEUI protocol cannot be routed and is best used in small local networks running relatively old operating systems from Microsoft and IBM:

· Microsoft Windows 3.1 or 3.11;

· Microsoft Windows 95;

· Microsoft Windows 98;

· Microsoft LAN Manager;

· Microsoft LAN Manager for UNIX;

· Microsoft Windows NT 3.51 or 4.0

· IBM LAN Server.

When migrating your network from Windows NT Server to Windows 2000 or Windows Server 2003, first configure servers and workstations that use NetBEUI to use TCP/IP. Although Windows 2000 systems support NetBEUI, Microsoft does not recommend using this protocol on later operating systems. However, if the network is small (less than 50 clients) and Internet access is not required, then the NetBEUI protocol may be more efficient than TCP/IP.

NetBEUIand reference modelOSI

The NetBEUI protocol corresponds to several layers of the OSI model. The physical and data link layers are used to interact between network interfaces. Within the Link Layer, LLC (Logical Link Control) and MAC (Media Access Control) sublayers are used to control the transmission of encoding and addressing of frames. The protocol also implements functions related to the Transport and Session layers (ensuring transmission reliability, acknowledging the receipt of packets, establishing and terminating sessions).

WhyNetBEUIworks well on networksMicrosoft

There are several reasons for answering the question posed in the section title. First, NetBEUI is easy to install because it does not need to be configured like other protocols (for example, TCP/IP requires an address, and IPX/SPX requires a frame type). Secondly, the protocol allows you to simultaneously support a large number of information exchange sessions on the network (up to 254 in earlier versions of the protocol, in previous versions this restriction has been removed). For example, according to Microsoft specifications, a Windows NT server can support 1000 sessions per network adapter (such tests were carried out for Windows 2000 servers). Thirdly, the NetBEUI protocol consumes little RAM and has high performance in small networks. Fourth, it implements reliable mechanisms for detecting and eliminating errors.

FlawsNetBEUI

The inability to route is the main disadvantage of the NetBEUI protocol in medium and large networks, including enterprise networks. Routers cannot forward a NetBEUI packet from one network to another because the NetBEUI frame does not contain information pointing to specific subnets. Another disadvantage of the protocol is that there are few network analyzers available for it (besides those tools that Microsoft has released).

Note

Practice 5-5 shows you how to install the NetBEUI protocol on a Windows 2000 computer.

ProtocolAppleTalk and systemMac OS

Apple has developed a family of protocols AppleTalk for organizing networks based on Macintosh computers running the Mac OS operating system. AppleTalk is a peer-to-peer network protocol, meaning it is designed to exchange data between Macintosh workstations even in the absence of a server. This fact is illustrated in Fig. 5.5, which shows how a switch is used to communicate between Macintosh computers. Novell NetWare, MS-DOS, Microsoft Windows operating systems can work with the AppleTalk protocol 9 x/ M.E. and Windows NT/2000/XP. The first version of the protocol was called AppleTalk Phase I and was released in 1983. In 1989, the still used version of AppleTalk Phase II was developed, which allows you to work a large number networked computers and provides interaction with large heterogeneous networks based on multiple protocols.

DIV_ADBLOCK27">

The maximum number of stations in the AppleTalk Phase I network is 254, and for the AppleTalk Phase II network this parameter is several million. Addressing in networks of the first type is carried out using node identification (ID), and in networks of the second type, both the node identifier and the network identifier are used when addressing. The final difference is that the AppleTalk Phase I protocol can only work on networks where there are no other protocols. The AppleTalk Phase II protocol operates on networks with multiple protocols (for example, IPX/SPX and TCP/IP).

Note

Although the AppleTalk protocol was designed as a peer-to-peer protocol, it can be used to exchange data between Mac OS X servers and Windows systems configured to work using this protocol.

ServicesAppleTalk

The AppleTalk protocol includes three basic services:

· remote access to network files using AppleShare File Server programs (in combination with the AppleTalk Filing Protocol);

· Print services based on AppleShare Print Server software (which use the Name Binding Protocol and Printer Access Protocol);

· file services based on AppleShare PC programs for DOS and Windows systems.

AppleTalkand reference modelOSI

In the AppleTalk stack, the original protocol lower level(according to the OSI model) is a protocol LocalTalk Link Access Protocol, LLAP, working on physical and Link levels and providing a legacy access method for data transfer. In this case, physical network interfaces, designed for the LocalTalk protocol, which can operate on small, slow networks with maximum quantity stations in the network equal to 32 (for a 300-meter segment with a bus topology). The permissible speed is 230.4 Kbps, which is extremely low for modern network technologies.

The LocalTalk network uses a process called contention to assign addresses. When the Macintosh computer is turned on, it competes with other computers for its address, resulting in a unique host identifier (ID). The next time the power is turned on, the computer may receive a different address.

Access MethodsAppleTalk

IN modern networks AppleTalk Phase II uses Ethernet or Token Ring access methods, and can use interfaces suitable for any other Ethernet or Token Ring devices. To simplify Ethernet communication, the AppleTalk stack includes a protocol EtherTalk Link Access Protocol, FLAP, operating at the Physical and Data Link levels. With its help, the CSMA/CD access method is implemented in AppleTalk networks with a bus or mixed topology (see chapter 2). Token ring networks use the protocol Token Talk Link Access Protocol, TLAP, also operating at the Physical and Link levels. It uses token passing and a ring/star topology (just like any other token ring network).

Network addressingAppleTalk

Addressing in AppleTalk networks using the ELAP and TLAP protocol is carried out using the protocol AppleTalk Address Resolution Protocol, AARP, which allows you to recognize the physical or MAC addresses of network adapters, so that these addresses can be inserted into AppleTalk frames. (If your Macintosh is configured for AppleTalk and IP, AARP is used to resolve physical and IP addresses.)

Protocols included in the stackAppleTalk

In addition to LLAP, ELAP, TLAP and AARP, there are other protocols that are part of the AppleTalk family. All of them are listed in table. 5.3.

Table 5.3. Protocols included in the stackApple

Abbreviatura	Full title	Description	LevelmodelsOSI
	AppleTalk Address Resolution Protocol	Used to recognize physical (MAC) addresses in Ethernet and Token Ring networks. If IP is used in addition to AppleTalk, AARP resolves computer and domain names to IP addresses	Channel and Network
	AppleTalk Data Stream Protocol	Provides guaranteed transmission of data streams at the receiving node	Session
	AppleTalk Filing Protocol	Allows workstations and servers to communicate with each other on Application level	Executive
	AppleTalk Session Protocol	Initiates, maintains and closes connections between stations. Determines the order in which data fragments are transmitted for reliable delivery to the receiving node	Session
	AppleTalk Transaction Protocol	Provides reliable data exchange between two nodes, for which each transaction is assigned a connection number	Transport
	Datagram Delivery Protocol	Used to deliver and route data between two communicating stations
	EtherTalk Link Access Protocol	Provides Ethernet communications using CSMA/CD access method in bus or mixed topologies	Physical and Channel
	LocalTalk Link Access Protocol	A legacy access method that controls communications at the Physical (via interfaces and cables) and Data Link layers in certain situations (for example, when contention for a unique ID occurs to provide addressing)	Physical and Channel
	Name Binding Protocol	Manages computer names and IP address registration, allowing clients to associate network services and processes with specific computer names	Transport
	Printer Access Protocol	Opens and closes communication sessions and provides network data transfer for print services	Session
	Routing Table Maintenance Protocol	Used to obtain network routing information when updating routing tables
	TokenTalk Link Access Protocol	Provides operation of token networks with ring/star topology	Physical and Channel
	Zone Information Protocol	Maintains a table of zones into which AppleTalk networks are divided and their corresponding routing tables	Session

CompatibilityAppleTalkWith systemsMac OS X,Windows 2000AndNetware

The native server platform for Macintosh computers is Mac OS X Server, which is based on the Mac OS X operating system. It allows you to share files and printers, manage network users and groups, and provide web services. Mac OS X and Mac OS X Server systems support both AppleTalk and TCP/IP.

A NetWare or Windows 2000 server can be used as a server for Macintosh computers if AppleTalk Phase II is available. For example, in order for a Windows 2000 server to be installed on a Macintosh computer network, the following components must be installed on it:

· AppleTalk Phase II;

· File Services for Macintosh;

· Print Services for Macintosh.

Once the AppleTalk protocol is installed, Windows 2000 Server will be able to communicate with Macintosh computers configured for AppleTalk Phase II. File Services for Macintosh allows you to allocate disk space on a Windows 2000 server on which Macintosh computers can store files using the AppleTalk protocol. Print Services for Macintosh allows Macintosh computers to access network printers supported by a Windows 2000 server.

Practice 5-6 will show you how to install the AppleTalk Phase II protocol on a Windows 2000 Server system, as well as File services Services for Macintosh and Print Services for Macintosh.

Note

The Mac OS X and Mac OS X Server operating systems are based on the UNIX kernel and even have a terminal window mode in which you can run numerous UNIX commands.

TCP/IP protocoland various server systems

Transmission Control Protocol/ Internet Protocol, TCP/ IP(Transmission Control Protocol/Internet Protocol) is the most common protocol stack currently used and is also the Internet Protocol. This section provides only a brief overview of TCP/IP in the context of a general understanding of the most important protocols. The TCP/IP stack is discussed in more detail in Chapter 6.

Most network server and workstation operating systems support TCP/IP, including NetWare servers, all Windows systems, UNIX, latest versions Mac OS, IBM's OpenMVS and z/OS systems, and DEC's OpenVMS. In addition, network equipment manufacturers create their own TCP/IP system software, including tools to improve device performance. The TCP/IP stack was originally used on UNIX systems and then quickly spread to many other types of networks.

Advantages of TCP/IP

Among the many benefits of the TCP/IP stack are the following:

· it is used in many networks and on the Internet, which makes it the international language of network communications;

· there are many network devices designed to work with this protocol;

· many modern computer operating systems use TCP/IP as the main protocol;

· For this protocol there are many diagnostic tools and analyzers;

· Many network specialists are familiar with the protocol and know how to use it.

Protocols and applications,included in the TCP/IP stack

In table 5.4 lists the protocols and applications included in the TCP/IP stack. Some of them have already been discussed earlier. A more detailed description is available in chapter b, and also in subsequent chapters.

Table 5.4. Protocols and applications included in the TCP/IP protocol stack

Abbreviation	Full title	Description	Model levelOSI
	Address Resolution Protocol	Provides resolution of IP addresses to MAC addresses	Channel and Network
	Domain Name System (application)	Maintains tables that associate computer IP addresses with their names	Transport
	File Transfer Protocol	Used to send and receive files	Session, Executive and Application
	Hypertext Transfer Protocol	Used to transmit data on the World Wide Web	Executive
	Internet Control Message Protocol	Used to generate network error reports, particularly when transmitting data through routers
	Internet Protocol	Controls logical addressing
	Network File System(application)	Used to transfer files over a network (designed for UNIX computers)	Session, Executive and Application
	Open Shortest Path First (protocol)	Used by routers to exchange information (routing data)
	Point-to-Point protocol	Used as a remote access protocol in combination with wide area network technologies
	Routing Information Protocol	Used when collecting routing data to update routing tables
	Remote Procedure Call (application)	Allows a remote computer to execute procedures on another computer (such as a server)	Session
	Serial Line Internet Protocol	Used as a remote access protocol in combination with wide area network technologies
	Simple Mail Transfer Protocol	Used to transmit email	Executive
	Transmission Control Protocol	Connection-oriented protocol that improves data transmission reliability	Transport
	Telecommunications Network (application)	Allows a workstation to emulate a terminal and connect to mainframes, Internet servers and routers	Session, Executive and Application
	User Data Protocol	Connectionless protocol; used as an alternative to TCP in cases where high reliability is not required	Transport

SNA protocol and IBM operating systems

Legacy IBM mainframes typically use stack protocols Systems Network Architecture, SNA, which was originally developed in 1974. In fact, SNA is a set of private protocols that use a token ring as an access method. Many details of the token networks created by IBM were subsequently included in the IEEE 802.5 standard. However, in an SNA network, the cable section is necessarily built on the basis of shielded twisted pair (STP), and the cables have strictly oriented markings (and wiring) (for example, one end of the cable must go to the mainframe, and the other to devices connected to the mainframe, such as controllers of disk drives or communication channels). This means that the SNA network also uses private (proprietary) cable connectors and network interfaces,

Protocol stackSNAand reference modelOSI

The SNA protocol stack is based on a seven-layer model (Table 5.5), reminiscent of the OSI reference model.

Table 5.5. Seven-level modelSNA

LevelSNA	Equivalent levelOSI	Purpose
Transaction Services	Applied	The highest level, controls the services on which application programs depend (for example, distributed databases data and applications running simultaneously on multiple mainframes)
Presentation Services	Representative	Controls data formatting and conversion (for example, conversion from ASCII to EBCDIC and vice versa), also performs data compression (although, unlike Representative OSI level, this level does not provide data encryption)
Data Flow Control	Session	Establishes and maintains communication channels between nodes, manages data flows and provides recovery from communication errors
Transmission Control	Transport	Ensures the reliability of data transmission from the source node to the receiving node, and also manages data encryption
Path Control		Manages routing and creation of virtual channels, fragments messages into blocks smaller sizes when transmitting data across heterogeneous networks (this task is performed by the OSI Transport Layer)
Data Link Control	Channel channel	Formats data into frames, provides marker access to the network for single-level data exchanges between computers
Physical Device Management (Physical Control)	Physical	Provides generation and encoding of electrical signals, operation of physical interfaces, network topology, and communication media (e.g. cable)

Advantages and Disadvantages of SNA

Like any protocol stack, SNA has both advantages and disadvantages. Noting the advantages, it should be said that the SNA architecture has existed for more than a quarter of a century and provides a reliable and proven means of exchanging data with IBM systems. A significant disadvantage is that SNA is a private (proprietary) protocol stack that requires special devices and additional training in configuration, management, and debugging procedures. For these reasons, SNA networks with IBM mainframes usually work very well, but it requires a large investment in staff training and network support.

Physical elements of an SNA network

In a traditional SNA network with IBM computers terminals are treated as physical modules of type 2 (type 2). A physical module is a device that can connect to or control access to the mainframe.

624 " style="width:467.8pt;border-collapse:collapse;border:none">

Abbrevia- tour orName

Full title

Description

LevelmodelsSNA

Advanced Peer-to-Peer Networking (Enhanced Peer-to-Peer Networking Protocol)

Provides peer-to-peer interactions between devices such as mainframes, minicomputers, gateways, and cluster controllers

Transmission Control

Customer Information Control System (subscriber information management system)

Data Flow Management and Representative Services

Distributed Data Management

Programs that provide remote access to information stored on IBM mainframes (for example, remote connection from another mainframe located at a distance)

Transaction Services

Information Management System (information management system)

Software environment, which provides programmers with basic tools for interacting with the SNA architecture (including secure access, file and drive management). An alternative to IMS is CICS

Data flow management Representative services

Network Control Program

Provides physical device addressing and additional logical addressing, as well as routing. Used for and management of SNA gateway communications (must be installed on any SNA gateway in order for workstations to access the mainframe through the gateway; see chapter 1 and 4, where gateways are discussed in more detail)

Channel Control and Route Control

Synchronous Data Link Control

Creates logical connections (virtual channels) in a network cable and coordinates data transfer over these connections, provides half-duplex and full-duplex communication in the channels

Physical Device Management and Channel Management

SNA Distributed Services

Software tools that control the transfer of documents. Used by email systems to transmit messages via specified addresses

Transaction Services

System Services Control Point system services)

Software VTAM control

Transmission Controls

Access method used by SNA networks

Physical device management Channel management

Virtual Telecommunications Access Method (virtual telecommunications access method)

Controls data transfer on an SNA network (for example, using flow control techniques). Provides digital data exchange

Transmission Control

DLC protocol for accessing IBM operating systems

If you are using Windows computers to access the mainframe running SNA 9 x, Windows NT and Windows 2000, then an alternative to the SNA gateway is to install the protocol Data Link Control, DLC. This protocol emulates SNA, and it can also be used to connect to some legacy network printers that can only work with it (for example, older Hewlett-Packard printers).

Advice

The DLC protocol is not supported on Windows XP. If you are considering upgrading to this system, please note that you will not be able to use the DLC to access IBM mainframes and may need an SNA gateway.

Basically, the DLC protocol is an alternative to TCP/IP in cases where some host uses SNA communications. The disadvantage of this protocol is that it is not routable. Additionally, it is not really designed for peer-to-peer communications between workstations, but only serves to connect to older IBM mainframes (eg ES9000) or IBM minicomputers (eg AS/400). Practice 5-7 shows how to install DLC on Windows 2000.

ProtocolDNAfor operating systemscomputersDigital (Compaq)

Architecture created in 1974 Digital Network Architecture (DNA) is the same age as SNA. DNA was used in the first networks of the Digital Equipment Corporation (DEC) and was otherwise called DECnet. Then this protocol stack was used much less frequently.

The DNA architecture provides for the use of Ethernet II frames (or DIX - an abbreviation for the names of the development companies Digital, Intel and Xerox) in a bus topology. One of the strengths of DNA is that from the very beginning, the architecture closely followed the OSI reference model. The disadvantage of DNA is that this architecture is private. In addition, following the acquisition of DEC by Compaq original computers DEC and DNA networks have become less popular. Even once-famous DEC Alpha-based computers are increasingly being replaced by Compaq-branded workstations and servers using Intel Itanium processors.

As DNA becomes less common in networks, the likelihood that you will encounter this architecture in practice decreases. However, for a general presentation in table. Section 5.7 lists some of the protocols and applications that make up the DNA stack.

Table 5.7. Protocols and applications included in the protocol stack

Abbreviation	Full title	Description	Model levelOSI
	Connectionless-Mode Network Service	Provides connectionless services (see chapter 2), as well as routing
	Connection Oriented Network Service	Provides connection-oriented services for routing and routing error control
	Digital Data Communications Message Protocol	Ensures that services operate with connection establishment and error control. At the level of electrical signals, it allows for half-duplex and full-duplex communication	Physical Channel (LLC sublayer)
	File Transfer, Access, and Management (file transfer, access and management)	Allows you to transfer files with text and binary content	Applied
	High-Level Data Link Control	Creates logical connections (virtual channels) in a network cable and coordinates data transfer between them. Controls the formatting of frames	Physical and Channel
		Complies with X.400 standard for postal services	Applied
	Naming Service	Provides network devices with naming services that translate a device's address into its name and vice versa (making it easier for users to work with devices)	Applied
	Network Virtual Terminal (network virtual terminal service)	Translates characters between Service terminals, DNA networks and host computers	Executive and Application

Improving the performance of local networks

The easiest way to improve network performance is to reduce the number of protocols sent through each router. This reduces the workload on routers, allowing them to process network traffic faster. With fewer protocols, there is also less unnecessary traffic generated on the network.

Issues for discussion

When choosing the protocols to use on your network, consider the following questions.

· Should packets be routed?

· What size is the network – small (less than 100 nodes), medium (100 – 500 nodes) or large (over 500 nodes)?

· What servers are used and what protocols do they require?

· Are there mainframes and what protocols do they require?

· Is there direct access to the Internet or connection to intranet applications using web technologies (virtual private network)?

· What speed is required for connections to the global network?

· Are there responsible applications?

If frames need to be routed (for example, on a corporate network), then the best protocol to use is TCP/IP, since it is routing-oriented and common in many networks. For small and medium-sized non-routable networks (less than 200 nodes) based on Windows NT servers and in the absence of an Internet connection, the NetBEUI protocol remains the best choice, providing fast and reliable communications. On NetWare networks (with servers earlier than 5.0), you can use IPX/SPX, although on a mixed network with older NetWare servers and newer Windows 2000 servers, you may need IPX/SPX and TCP/IP protocols. The NWLink protocol is good remedy for connecting Windows 9x/NT/2000 systems to older NetWare servers.

Communication channel problem

Having a connection to the Internet or web services requires TCP/IP to be deployed, and FTP services can be used to transfer files. TCP/IP is also best used for communications with temporary mainframes and UNIX computers, since connecting to a mainframe or to an application running on a UNIX computer may require Telnet terminal emulation. You can also use the DLC protocol to connect to IBM mainframes and minicomputers (if they are running in an SNA environment). Finally, DNA protocol may still be needed on a network containing older DEC computers (eg DEC VAX).

Note

TCP/IP is the best protocol for medium and large networks. It is routable, robust for mission-critical applications, and has a robust error control mechanism. In such networks, it is important to have network monitoring and fault analysis tools. As stated in chapter 6, the TCP/IP stack has the protocols necessary to solve such problems.

In many cases, different network applications require different LAN protocols. Sometimes in modern networks, TCP/IP, NetBEUI, IPX/SPX, SM and even DNA protocols are used in any combination. As you already know, the protocols deployed are related to the type of operating systems used. Their choice is also influenced by the availability of connections to global networks (for example, to access the Internet you need the TCP/IP protocol, which may also be required to connect local networks to each other via a global network). If, say, TCP/IP is used by servers on one LAN, and workstations on another network must access those servers, then both LANs and the connecting WAN must support TCP/IP protocol transmission.

Removing unnecessary protocols

Sometimes workstations on a network remain configured to use multiple protocols even after all hosts and servers have been converted to TCP/IP. In this case, you can easily improve network performance by removing unnecessary protocols from workstations. Practice Exercise 5-8 teaches you how to remove DLC from Windows 2000, and Practice 5-9 teaches you how to remove Client Service for NetWare (and NWLink IPX/SPX) from Windows 2000 and Windows XP Professional.

Summary

· To a large extent, the architecture of networks is determined by protocols. Many networks use multiple protocols to access the various operating systems of network servers and host computers.

· Typically used LAN protocols are determined by the type of network server operating system used in specific network. One of the oldest network systems is NetWare, which works with the IPX/SPX protocol stack and provides data transfer between older versions of NetWare servers and workstations (as well as other servers) connected to the servers. The IPX/SPX protocol is implemented in thousands of local networks, since NetWare is one of the common network operating systems. However, nowadays, due to the fact that many networks are connected to the Internet, new versions of NetWare (5.0 and higher) are focused on working with the more universal TCP/IP protocol stack.

· The native protocol for Windows NT Server systems is NetBEUI, the emergence of which is associated with the development of the LAN Manager network operating system, which Microsoft began jointly with IBM. Medium and large networks with Windows NT servers often use the TCP/IP stack. With the advent of Windows 2000 and Windows Server 2003, the TCP/IP protocol replaced NetBEUI, which is determined by the requirements of the service Active Directory and the need for Internet access.

· AppleTalk is a protocol used by Macintosh computers running the Mac OS and Mac OS Server operating systems. Windows NT, Windows 2000, Windows Server 2003, and Novell NetWare also support AppleTalk.

· Some network server operating systems (in particular, UNIX) were initially designed to work with the TCP/IP stack (as well as the Internet). Other network operating systems (such as NetWare, Windows NT, and Mac OS Server) implemented the TCP/IP stack after those systems were created.

· First IBM systems the SNA protocol stack was used, which provided data exchange between mainframes (mini-computers) and terminals, controllers and printers, as well as between different computers. Windows operating systems have the ability to install the DLC protocol to emulate SNA communications.

· The DNA protocol stack was designed for use on DEC computer-based networks, but is rarely used today as the number of such computers on networks has decreased significantly.

· Simple and effective way Improving the performance of a local network is a periodic analysis of the protocols used and the removal of those protocols that are no longer used. For access to computers and printers.

· Until the early 1990s network technologies were primarily broken down in the area of ​​local network protocols. Currently, the architecture of these protocols has found its logical conclusion in the TCP/IP stack, and private protocols (such as IPX/SPX and NetBEUI) are used less frequently.

Surely many people understand that computers located in various points planets, increasing in number, sooner or later had to “learn” to communicate with each other and become capable of working together. The means of such communication have become local and global. As for local networks, these are networks that connect computers located on short distances from each other, for example, in the same building. The main purpose of global networks is to connect networks and computers separated by vast distances - hundreds and thousands of kilometers. The largest network on the planet is the Internet.

Even understanding theoretical foundations functioning of networks, often does not give a person the opportunity to continue a rather simple phrase: a computer network protocol is... Below we will try to figure out what they are and what they are needed for. Let us immediately note that the computer network protocol is a fundamental point that makes it possible to organize communication between computers, regardless of the distance that separates them.

The fact is that easy connection one computer to another - the step required to create a computer network is not enough. In order for the ability to transmit information on a network to become accessible, it is necessary that computers “understand” each other. A computer network protocol is a specially designed means by which computers “communicate” through a network in a “language” that is understandable to each other. In addition, a computer network protocol is a set of rules, following which it is possible to organize between computers.

In order to fully understand what a protocol is, let’s digress from the computer industry. Even a person who has never encountered networks and the Internet has encountered devices in everyday life, the functioning of which is also based on specially developed protocols. For example, ordinary telephone communication, which everyone uses, is based on its own protocol, which allows devices to establish the fact that the handset on the device that is receiving the call has been picked up, to recognize the fact of disconnection, as well as the caller’s number.

We hope it is now clear why the computer world needed a single language (called a protocol) that every computer in the world could understand.

The main Internet protocols are TCP/IP, POP3, SMTP, FTP, HTTP, IMAP4, WAIS, Gorpher, WAP. Each of these protocols performs specific functions.

The basic protocol on the Internet is TCP/IP - a protocol created by the best minds of humanity. One of the creators of this protocol was Vinton Cerf. Today this man is considered the “father of the Global Network.” Now Vinton Cerf works as senior vice president for the creation and development of Internet architecture at MCI WorldCom Inc.

Work on the Transmission Control protocol (that is, TCP/IP) was completed in 1972 by a group of developers led by Vinton Cerf.

Initially, TCP/IP was developed for the needs of the US Department of Defense, but then the protocol outgrew its purpose and became a basic set of rules that allowed the global Internet to develop rapidly. In addition, small networks using intranets operate using this protocol. Today, TCP/IP standards are open protocols and are constantly being improved.

It is worth noting that TCP/IP is not actually one protocol; at its core, it is a whole list of protocols that work together. The protocol consists of two levels. Purpose of the protocol top level- TCP is the organization of the correct transformation of data into information packets, which, upon reaching the receiving party, become the basis for constructing the original message. The developers assigned the lower-level protocol, IP, the responsibility to monitor the correct delivery of messages to the destination address.