Что такое multilink ppp

Multilink PPP on Back-to-Back Routers with Multiple Serial Interfaces

Available Languages

Download Options

Contents

Introduction

In some environments, it may be necessary to bundle serial links to act as single aggregate bandwidth. This document describes how to configure a Cisco 2503 Access Server to bundle two serial interfaces with two different methods:

These configurations can be used for routers connected by leased lines or routers that have the channel service unit or data service unit (CSU/DSU) or ISDN terminal adapter (TA) configured to dial. (Cisco routers have not been configured to dial telephone numbers.) You can add additional features to this configuration to suit your needs.

Prerequisites

Requirements

There are no specific requirements for this document.

Components Used

The information in this document is based on these software and hardware versions:

Cisco 2503 routers

Cisco IOS® Software Release 12.2(7b)

The information in this document was created from the devices in a specific lab environment. All of the devices used in this document started with a cleared (default) configuration. If your network is live, make sure that you understand the potential impact of any command.

Related Products

This configuration can be used with any two routers with two WAN serial interfaces each. You can use WIC-1T, WIC-2T, or fixed WAN serial interfaces.

Conventions

For more information on document conventions, refer to Cisco Technical Tips Conventions.

Configure

In this section, you are presented with the information to configure the features described in this document.

Network Diagram

This document uses this network setup:

Configurations

This configuration was tested with Cisco IOS Software Release 12.2(7b) on Cisco 2500 series routers. The same configuration concepts would apply to a similar router topology or other Cisco IOS Software releases.

Example 1: Interface Dialer

Example 1 Notes

The ppp direction callout command is a hidden command used when a router is confused as to who dialed who (when connected back-to-back or connected by leased lines and the CSU/DSU or ISDN TA are configured to dial). The ppp direction callin command may also be used. Use either of these commands.

For a local router, use ppp direction callout.

For a remote router, use ppp direction callin.

If you do not use this command, the router connects for a moment and then disconnects.

The above configuration describes two routers, Router1 and Router2, which have two serial interfaces to be bundled to act as a single aggregate bandwidth. On both routers, interface loopback is configured and integrated with interface dialer using ip unnumbered loopback0. Both serial interfaces are configured for no IP address. Physical interfaces Serial0 and Serial1 are configured with dialer rotary-group and associated with a single logical interface dialer.

Dialer-list 1 defines the interesting traffic, which in turn triggers the dialer-group 1 command on interface dialer 1 to bring up the link. The dialer map statement defined on both routers maps the interface dialer 1 to the peer router IP address, and the hostname defined serves in CHAP authentication. The static route defined routes the traffic to the destination.

The ppp authentication chap command enables the PPP negotiation. The dialer load-threshold command sets a load that triggers to bring the second serial line. The ppp multilink and dialer rotary-group commands have been configured so that both serial interfaces can be bundled together as one Virtual-Access interface for aggregate bandwidth. The ppp direction callout command determines which side has to make a callout during PPP negotiation and CHAP authentication.

Verify Example 1

This section provides information you can use to confirm your configuration is working properly.

Troubleshoot Example 1

This section provides information you can use to troubleshoot your configuration.

debug Commands

On Router1, the debug ppp authentication command shows the success of CHAP.

On Router2, the debug ppp authentication command shows «Waiting for peer. «

Example 2: Virtual Template

The configuration above describes the virtual template configured on Router1 and Router2. In this example, both routers are configured with virtual templates. The routers are connected back-to-back, and the multilink session does not idle out. No static routes are needed: a host route is installed after PPP negotiations.

Use Cisco IOS Software Release 11.3 or later to use virtual template for PPP multilink.

The virtual template command dynamically creates a virtual access interface and applies them to physical serial interfaces with the multilink virtual-template command. The parameters (such as ppp authentication chap) configured in interface virtual template are applied for both serial interfaces. The ppp multilink command in the interface virtual-template bundles the physical serial interfaces to form a virtual-access in order to aggregate the bandwidth.

Verify Example 2

This section provides information you can use to confirm your configuration is working properly.

These commands may also help:

show ip route connected—To see if the IP route for the virtual-access is installed.

show interface virtual-access x—To check the status of a particular virtual-access interface. In the example above, the virtual-access interface number is 1.

Troubleshoot Example 2

This section provides information you can use to troubleshoot your configuration.

Configure timestamps in the global configuration as follows:

Use these commands to troubleshoot:

debug ppp negotiation—To see if a client passes PPP negotiation. You can also check which options (callback, Multilink PPP [MLP], and so on) and which protocols (IP, IPX, and so on) are negotiated.

debug ppp authentication—To see if a client passes authentication.

debug vtemplate—To see what virtual-template configurations are used.

debug vprofile—To see what configuration options are applied to the virtual-access interface.

Источник

PPP MultiLink

Sometimes we see a network design where we have more than one serial link between two routers. Perhaps a single serial link doesn’t provide enough bandwidth or you want some extra redundancy. Each interface on a router requires a different IP address, so with two serial links you might end up with a design that looks like this:

Above we see that the first link uses the 192.168.12.0/24 subnet, the second link uses 192.168.21.0/24.When we use two subnets between two routers and when you configure a routing protocol like OSPF or EIGRP, it will form two neighbor adjacencies:

When the metric on both links is the same, our routing protocol will use load balancing so both links will be used. By default, Cisco IOS will load balance based on the destination. For example, if 2.2.2.2 would be behind R2 then R1 might use Serial 0/0/0 for all traffic destined to 2.2.2.2. Another destination might use the Serial 0/1/0 interface.

What if we add more serial links? For each serial link, we will have to configure another subnet and we will get another neighbor adjacency. It will work but we can make things much easier…

Multilink PPP allows us to combine multiple physical serial links into a single logical link. We will only need a single subnet for the logical multilink interface and if you use OSPF or EIGRP, only a single neighbor adjacency is required:

MLPPP also does load balancing on layer two. This is pretty efficient, when it receives some data like an IP packet that has to be forwarded over the multilink interface, it will fragment the IP packet in two pieces and forwards it on the multilink interface. The receiving router takes the fragments and reassembles them into the original IP packet:

By using PPP multilink, we simplify our configuration on layer three since there will be only one logical interface to work with.

Configuration

Let’s see how we configure PPP multilink. I will use the following topology:

First, don’t forget to specify a clock rate. I will check which of my routers has the DCE end of the cable:

R2 is the DCE and you can see a clock rate was configured. Let’s enable PPP encapsulation on all serial interfaces on R1:

And we do the same thing on R2:

Now we can create the multilink interface on both routers:

You can pick whatever multilink interface number you like, just make sure that it’s the same on both ends. This is where we configure the IP address.

On the physical serial interfaces, we have to configure to which multilink interface they belong with the ppp multilink group command. Let’s start with R1:

And let’s do the same thing on R2:

That takes care of the PPP multilink configuration.

To see the PPP multilink interface in action, let’s configure OSPF on both routers:

That completes our configuration.

Verification

Let’s verify our work step-by-step. First, let me quickly show you multilink interface in the running configuration:

We only created the multilink 1 interface and configured an IP address but you can see that Cisco IOS automatically added the ppp multilink and ppp multilink group commands.

Let’s continue, I want to make sure that the physical interfaces are up and running:

The interfaces are up/up which tells us that the clock rate is configured correctly and that PPP encapsulation is configured on both ends. Let’s take a look at the multilink interface:

The multilink interface is up too, it also shows me the IP address. Let’s take a closer look at the PPP details of the multilink interface:

The output above tells me that the multilink interface is up and has bundled Serial 0/0/0 and Serial 0/1/0 to the multilink interface. One more command to verify the PPP multilink interface, let’s look at its IP settings:

Источник

Available Languages

Download Options

Contents

Introduction

Multilink PPP (also referred to as MP, MPPP, MLP, or Multilink) provides a method for spreading traffic across multiple physical WAN links while providing packet fragmentation and reassembly, proper sequencing, multivendor interoperability, and load balancing on inbound and outbound traffic.

MPPP allows packets to be fragmented. These fragments are sent simultaneously over multiple point-to-point links to the same remote address. The multiple physical links come up in response to a user-defined load threshold. This load can be measured on just inbound traffic, on just outbound traffic, or on either; however, it cannot be measured on the combined load of both inbound and outbound traffic.

For dial connections, MPPP can be configured for ISDN Basic Rate Interfaces (BRIs) and Primary Rate Interfaces (PRIs), as well as for asynchronous serial interfaces. It can also be configured for non-dial serial interfaces, though this functionality is not specifically addressed in this document. This document will address configuration of basic MPPP for Dial-on-Demand Routing (DDR). Multichassis Multilink PPP will not be covered in this document; see the Multichassis Multilink PPP (MMP) documentation for more information.

Before You Begin

Conventions

For more information on document conventions, see the Cisco Technical Tips Conventions.

Prerequisites

There are no specific prerequisites for this document.

Components Used

The information in this document is based on the software and hardware versions below.

Multilink PPP was first introduced in Cisco IOS® Software Release 11.0(3)

Cisco IOS Software Release 11.3 was used in this example.

The information presented in this document was created from devices in a specific lab environment. All of the devices used in this document started with a cleared (default) configuration. If you are working in a live network, ensure that you understand the potential impact of any command before using it.

What Multilink PPP Does

Traffic routed across an MPPP link via its controlling interface (a Virtual Access interface) will be fragmented, with the fragments being sent across the different physical links. At the remote end of the link, the fragments are reassembled and forwarded to the next hop toward their ultimate destination.

Configuring Multilink PPP

This section addresses the commands and the different methods of configuring MPPP on a router.

Commands

Note: If you add this command, you must disconnect any existing connections and then reconnect for the new multilink parameters to be applied. Because multilink is negotiated during the call setup, any changes to multilink are not implemented on connections that have completed the link control protocol (LCP) negotiation.

Tip: Often, customers will configure the command dialer load-threshold 1 because they want all of their B-channels to be used immediately for every call. The theory behind this is that if all the B- channels go up at once and the entire ISDN pipe is used for each call, the call should be shorter in duration because it will take less time to transfer the user data.

Legacy DDR

This section addresses how to configure Multilink PPP using Legacy DDR (rotary-group and dialer maps).

Because ISDN interfaces are considered to be «Dialer» interfaces, few commands are required to make an ISDN interface capable of making MPPP connections. For example, it is not necessary to configure a dialer rotary group unless you are using more than one BRI or PRI.

Following is an example of a BRI configured to make a simple dial-on-demand PPP connection:

Only two commands must be added to this interface’s configuration to make MPPP possible. The router at the other end of the call must be similarly configured. These two commands are:

In circumstances where two or more physical interfaces need to be bundled together (for example, when using async or serial interfaces, or more than one ISDN interface) a different method must be used. In these cases, a dialer rotary group must be configured and a Dialer interface must be added to the configuration of the router in order to control the MPPP connection. In brief, a «logical» interface must control the «physical» interfaces.

In order to accomplish this, you must:

Place the physical interfaces into a rotary group.

Create a logical («Dialer») interface as the lead for the rotary group.

Configure the Dialer interface to do MPPP.

Follow these steps to configure MPPP on multiple interfaces:

Put the physical interfaces into a rotary group by using the dialer rotary-group number command. In this example, the asynchronous interface is put into rotary group 1:

Note: Be sure to use the no shutdown interface configuration command if the router has never been configured or if the router has been set back to its default configuration.

To create a Dialer interface, use the interface dialer number global configuration command. In this example, interface Dialer 1 is created:

Note: The number argument of the interface dialer command must be the same as the number of the rotary group configured in Step 1.

Use the show running-config command to see the default configuration of a dialer interface:

Next, configure the Dialer interface in order to place or receive calls. The essential commands for MPPP are the same as in Step 1:

For examples of complete DDR configurations with MPPP, see the PPP Support Page

Dialer Profiles

Configuring Multilink PPP on Dialer Profiles is similar to that for Legacy DDR. The ppp multilink command must be configured on both the physical interface and dialer interface. The dialer load-threshold command should be configured on the Dialer interface. For example,

For more information on dialer profiles refer to the document Configuring and Troubleshooting Dialer Profiles

Verify MPPP Operation

In order to verify the proper operation of an MPPP connection, use the debug ppp negotiation command. The critical elements that must be negotiated in the LCP phase are the Maximum Receive Reconstructed Unit (MRRU) and the Endpoint Discriminator (EndpointDisc):

As with the other elements of LCP negotiation, the MRRU and EndpointDisc must be agreed to by both ends of the connection during the exchange of CONFREQs and CONFACKs. Both ends of the connection must send CONFACKs for the protocol to be established. For more information on how to read debug ppp negotiation output refer to the document Understanding debug ppp negotiation Output.

After MPPP has been successfully negotiated during the LCP phase of PPP negotiation and Challenge Handshake Authentication Protocol (CHAP) or Password Authentication Protocol (PAP) have completed successfully, a Virtual Access interface will be created by the Cisco IOS Software to represent the MPPP bundle. For more information on the uses and theory behind Virtual Access interfaces, please see the Virtual Access PPP Features in Cisco IOS documentation.

The creation of the Virtual Access interface is signaled in the debug ppp negotiation output by the following:

From this point forward, PPP negotiation of the Network Control Protocols is handled by the Virtual Access interface. For example:

Once the MPPP connection has been established, information on the connection can be found in the output of the show ppp multilink command:

In addition, the show interface command is valid for the Virtual Access interface as it is for any other physical or logical interface. The same type of information will be presented as would appear in any other show interface output.

Источник

Двухточечный протокол

Две производные от PPP, двухточечный протокол через Ethernet (PPPoE) и двухточечный протокол через ATM (PPPoA), чаще всего используются интернет-провайдерами для установления соединения с клиентами через цифровую абонентскую линию (DSL).

СОДЕРЖАНИЕ

Описание [ править ]

Автоматическая самоконфигурация [ править ]

RFC 1994 описывает протокол аутентификации с вызовом и рукопожатием (CHAP), который предпочтительнее для установления коммутируемых соединений с интернет-провайдерами. Несмотря на то, что протокол аутентификации паролей (PAP) считается устаревшим, он все еще иногда используется.

После того, как соединение установлено, может иметь место дополнительная конфигурация сети ( уровень 3 ). Чаще всего используется протокол управления интернет-протоколом (IPCP), хотя когда-то были популярны протокол управления межсетевым обменом пакетами (IPXCP) и протокол управления AppleTalk (ATCP). [ необходима цитата ] Протокол управления Интернет-протоколом версии 6 (IPv6CP) будет широко использоваться в будущем, когда IPv6 заменит IPv4 в качестве доминирующего протокола уровня 3.

Несколько протоколов сетевого уровня [ править ]

PPP позволяет нескольким протоколам сетевого уровня работать на одном и том же канале связи. Для каждого используемого протокола сетевого уровня предоставляется отдельный протокол управления сетью (NCP), чтобы инкапсулировать и согласовывать варианты для нескольких протоколов сетевого уровня. Он согласовывает информацию сетевого уровня, например сетевой адрес или параметры сжатия, после того, как соединение установлено.

Например, Интернет-протокол (IP) использует протокол управления IP ( IPCP ), а межсетевой обмен пакетами (IPX) использует протокол управления Novell IPX ( IPX / SPX ). NCP включают поля, содержащие стандартизованные коды, чтобы указать тип протокола сетевого уровня, который инкапсулирует PPP-соединение.

Следующие NCP могут использоваться с PPP:

Обнаружение зацикленных ссылок [ править ]

Параметры конфигурации [ править ]

В предыдущем разделе было представлено использование опций LCP для удовлетворения конкретных требований к подключению к глобальной сети. PPP может включать следующие параметры LCP:

Кадр PPP [ править ]

Структура [ править ]

Если оба одноранговых узла соглашаются на сжатие поля адреса и поля управления во время LCP, эти поля пропускаются. Аналогично, если оба одноранговых узла соглашаются на сжатие поля протокола, то байт 0x00 может быть опущен.

Инкапсуляция [ править ]

Поле Flag присутствует, когда используется PPP с кадрированием, подобным HDLC.

FCS вычисляется по полям Address, Control, Protocol, Information и Padding после инкапсуляции сообщения.

Активация линии и фазы [ править ]

По нескольким ссылкам [ править ]

Многоканальный PPP [ править ]

В одну линию PPP кадры не могут поступать не по порядку, но это возможно, когда кадры разделены между несколькими соединениями PPP. Следовательно, Multilink PPP должен пронумеровать фрагменты, чтобы их можно было снова расположить в правильном порядке по прибытии.

Мультиклассовый PPP [ править ]

С помощью PPP невозможно установить несколько одновременных отдельных соединений PPP по одному каналу.

Это также невозможно с Multilink PPP. Multilink PPP использует смежные номера для всех фрагментов пакета, и, как следствие, невозможно приостановить отправку последовательности фрагментов одного пакета, чтобы отправить другой пакет. Это предотвращает многократный запуск Multilink PPP по одним и тем же каналам.

Туннели [ править ]

Производные протоколы [ править ]

Как протокол уровня 2 между обоими концами туннеля [ править ]

Поскольку в туннеле есть только две конечные точки, туннель является двухточечным соединением, и PPP является естественным выбором в качестве протокола уровня канала данных между виртуальными сетевыми интерфейсами. PPP может назначать IP-адреса этим виртуальным интерфейсам, и эти IP-адреса могут использоваться, например, для маршрутизации между сетями по обе стороны туннеля.

IPsec в режиме туннелирования не создает виртуальные физические интерфейсы в конце туннеля, поскольку туннель обрабатывается непосредственно стеком TCP / IP. L2TP может использоваться для предоставления этих интерфейсов, этот метод называется L2TP / IPsec. В этом случае PPP также предоставляет IP-адреса концам туннеля.

Стандарты IETF [ править ]

Источник