Operation Manual MPLS. Table of Contents

Transcription

1 Table of Contents Table of Contents Chapter 1 MPLS Architecture MPLS Overview MPLS Basic Concepts FEC Label LDP MPLS Architecture MPLS Network Structure LSP Establishment LSP Tunnel and Hierarchy Forwarding Labeled Packets LDP Overview LDP Basic Concepts LDP Working Process LDP Basic LDP Loop Detection Constrain-based Routing LDP MPLS and Other Protocols MPLS and Routing Protocols MPLS Extension by RSVP MPLS Application MPLS-Based VPN MPLS-Based Traffic Engineering MPLS-Based QoS Introduction to MPLS Configuration MPLS Basic Capability Configuration Defining MPLS LSR ID Entering MPLS View Configuring the Topology-Driven LSP Setup Policy Configuring Static LSP Configuring IP TTL Duplication of MPLS Configuring MPLS to Return ICMP Responses by IP routing LDP Configuration Enabling/Disabling LDP Enabling/Disabling LDP on Interface Configure LDP extended discovery mode i

2 Table of Contents Configuring Session Parameters Configuring Remote LDP Peer Session Parameters LDP Loop Detection Control Configuring LDP Authentication Mode Displaying and Debugging MPLS Displaying and Debugging MPLS Displaying and Debugging LDP MPLS Configuration Example Troubleshooting MPLS Configuration ii

3 Chapter 1 MPLS Architecture Chapter 1 MPLS Architecture 1.1 MPLS Overview MPLS (Multiprotocol Label Switching) encapsulates packets with labels of short and fixed length. MPLS obtains service from various link layers (such as PPP, ATM, Frame Relay, and Ethernet) and provides connection-oriented service for network layer. MPLS can obtain support from IP routing protocol and control protocol and, at the same time, it supports policy-based restraint route. It possesses powerful and flexible routing functions and is capable of satisfying the requirements for the network from various new applications. This technology was initially originated from IPv4. However, its core technology can be extended to multiple network protocols (IPv6, IPX, etc). MPLS is a protocol initially developed for increasing forwarding speed of routers. However, it has gained wider applications in traffic engineering, VPN, and QoS and is becoming an important standard for large-scale IP networks. 1.2 MPLS Basic Concepts FEC Forwarding Equivalent Class (FEC), in fact, is a kind of classify-and-forward technology. It categorizes packets with the same forwarding strategy (same destination addresses, same forwarding routes and same QoS levels) into one group, which is called a FEC. The FEC classification is based on the network layer address. Packets in the same FEC will be processed in absolutely the same way Label I. Label definition Label is a locally significant and short identifier with fixed length, which is used to identify a particular FEC. When reaching at MPLS network ingress, packets are divided into different FECs, on which different labels are encapsulated. Later forwarding is based on these labels. II. Label structure The structure of the label is shown in Figure 1-1. Label Exp S TTL Figure 1-1 Label structure 1-1

4 Chapter 1 MPLS Architecture Label is located between the link layer header and the network layer packet, with the length of 4 bytes. A label contains four fields: Label: label value, 20bits, used as the pointer for forwarding. Exp: 3bits, reserved for test. S: 1bit, MPLS supports hierarchical label structure, i.e., multi-layer label. Value 1 refers to the label of bottom layer. TTL: 8 bits, similar to TTL in IP packets. III. Label operations 1) Label mapping There are two types of label mapping: one is label mapping at ingress routers and the other is label mapping in MPLS domain. The first type is also called ingress LSR, in which input packets are grouped on a certain principle into multiple FECs. The corresponding labels are mapped to these FECs and the mapping results are recorded into label information base (LIB). In simple words, it is to assign a label to a FEC. The second type is also called incoming label mapping (ILM), that is, to map each input label to a series of next hop label forwarding entries (NHLFE). The packets are forwarded along the paths based on the mapping results. 2) Label encapsulation Label encapsulation in different media is illustrated in Figure 1-2. Ethernet/SONET/ SDH packets Ethernet header /PPPheader Label L3 data Frame-mode ATM packets ATM header Label L3 data Cell-mode ATM packets VPI/VCI L3 data Figure 1-2 Label position in packet 3) Label assignment and distribution Label distribution refers to the process for a FEC to create corresponding label switching path (LSP). In the MPLS architecture, the decision to bind a particular label to a particular FEC is made by downstream LSR, and the downstream LSR notifies the upstream LSR. That is to say, the label is specified by the downstream LSR, and the label is distributed from downstream to upstream. 1-2

5 Chapter 1 MPLS Architecture Two label distribution modes are available in MPLS: Downstream unsolicited (DU) mode and downstream on demand (DoD) mode. For a specific FEC, if LSR originates label assignment and distribution even without receiving label request messages from upstream, it is in DU mode. For a specific FEC, if LSR begins label assignment and distribution only after receiving label request messages from upstream, it is in DoD mode. The upstream and downstream, which have adjacency relation in label distribution, should reach agreement on label distribution mode. For the peers, LSR can use LDP messages to distribute labels or bear labels over other routing protocol messages. Note: Upstream and downstream are just on a relative basis: For a packet forwarding process, the transmit router serves as upstream LSR and receive router serves as downstream LSR. 4) Label control mode There are two types of label control modes: independent mode and ordered mode. In independent control mode, each LSR can notify label mapping messages anytime. In ordered control mode, a LSR can send label mapping messages to upstream only when it receives a specific label mapping messages of the next hop of a FEC or the LSR serves as LSP (Label distribution protocol) egress node. 5) Label retention mode There are two label-retention modes: liberal label retention mode and conservative label retention mode. For a specific FEC, if LSR Ru has received the label binding from LSR Rd, in case Rd is not the next hop of Ru and Ru saves this binding, then it is called liberal label retention mode. If Ru discards this binding, then it is called conservative label retention mode. In case it is required that LSR is capable of adapting route variation rapidly, liberal label retention mode can be adopted. In case it is required that a few labels are saved in LSR, then conservative label retention mode can be used LDP Label distribution protocol (LDP) is the signaling control protocol in MPLS, which controls binding of exchange labels and FECs between LSRs and coordinates a series of procedures between LSRs. 1-3

6 Chapter 1 MPLS Architecture 1.3 MPLS Architecture MPLS Network Structure The basic composing unit of MPLS network is LSR (Label Switching Router). It runs MPLS control protocol and L3 routing protocol, exchanges routing messages with other LSRs and create the routing table, maps FECs with IP packet headers, binds FECs with labels, distributes label binding messages, establishes and maintains label forwarding table. The network consisting of LSRs is called MPLS domain. The LSR that is located at the edge of the domain edge LSR (LER, Labeled Edge Router). It connects an MPLS domain with a non-mpls domain or with another MPLS domain, classifies packets, distributes labels (as egress LER) and distracts labels. The ingress LER is termed as ingress and egress LER as egress. The LSR located inside the domain is called core LSR. The core LSR can be either the router that supports MPLS or the ATM-LSR upgraded from ATM switch. It achieves label swapping and label distribution. The labeled packets are transmitted along the LSP (Label Switched Path) composed of a series of LSRs. Label Switched Path (LSP) Ingress Egress MPLS Core LSR MPLS Edge LSR (LER) Figure 1-3 MPLS basic principle LSP Establishment Actually, LSP establishment refers to the process of binding FEC with the label, and then advertising this binding to the adjacent LSR on LSP. This process is implemented via Label Distribution Protocol (LDP). LDP regulates the message in interactive processing and message structure between LSRs as well as routing mode. Refer to the following section for the detail description of LDP 1-4

7 Chapter 1 MPLS Architecture LSP Tunnel and Hierarchy I. LSP tunnel MPLS supports LSP tunnel technology. On an LSP path, LSR Ru and LSR Rd are upstream and downstream for each other. However, the path between LSR Ru and LSR Rd may not be part of the path provided by routing protocol. MPLS allows establishing a new LSP path <Ru R1...Rn Rd> between LSR Ru and LSR Rd, and LSR Ru and LSR Rd are respectively the starting point and ending point of this LSP. The LSP between LSR Ru and LSR Rd is referred to as the LSP tunnel, which avoids the traditional encapsulated tunnel on the network layer. If the route along which the tunnel passes and the route obtained hop by hop from routing protocol is in consistent, this tunnel is called hop-by-hop routing tunnel. And if the two routes are not in consistent, then the tunnel of this type is called explicit routing tunnel. R1 R2 R3 R4 Layer 1 R21 Figure 1-4 LSP tunnel R22 Layer 2 As shown in Figure 1-4, LSP <R2 R21 R22 R3> is a tunnel between R2 and R3. II. Multi-layer label stack When the packet is sent in LSP tunnel, there will be multiple layers for the label of the packet. Then, on the ingress and egress of each tunnel, it is necessary to implement incoming and outgoing operation for the label stack. For each incoming operation, the label will be added with one layer. And there is no depth limitation for the label stack from MPLS. The labels are organized according to the principle last in first out in the label stack, and MPLS processes the labels beginning from the top of the stack. Suppose that a packet has the label stack depth of m, then the label at the bottom of the stack is the label of first level, and the label at the top of the stack is the label of level m. The packet with no label can be regarded as the packet of blank label stack (namely, the label stack depth is zero) Forwarding Labeled Packets In Ingress, the packets entering the network are classified into Forwarding Equivalence Class (FEC) according to their characteristics. Usually, FEC is classified according to the IP address prefix or host address. The packets with the same FEC will pass through the same path (i.e., LSP) in MPLS area. LSR assigns a short label of fixed length for the incoming FEC packet, and then forwards it through the corresponding interface. 1-5

8 Chapter 1 MPLS Architecture On the LSR along the LSP, the mapping table of the import/export labels has been established (the element of this table is referred to as Next Hop Label Forwarding Entry (NHLFE)). When the labeled packet arrives, LSR only needs to find the corresponding NHLFE from the table according to the label and replace the original label with the new special label, and then forwards the labeled packet. This process is called Incoming Label Map (ILM). On the Ingress, MPLS specifies FEC of specific packet, and the following routers only need to forward it by label switching therefore this method is much simpler than the routine network layer forwarding. Note: TTL Processing: For labeled packet, it is necessary to copy the TTL value in the original IP packet into the TTL field in the label. While forwarding the label type packet, LSR will perform minus one operation for the TTL field of the label on the top of the stack. When the label is out of the stack, the TTL value on the top of the stack is copied back to IP packet or the label of lower layer. However, while LSP goes through the non-ttl LSP segment composing of ATM-LSR or FR-LSR, the LSR inside the non-ttl LSP segment is not capable of processing TTL field. In this case, it is necessary to carry out unified processing for TTL while entering non-ttl LSP segment, namely, to reduce for one time the value that reflects the length of this non-ttl LSP. 1.4 LDP Overview The LDP is responsible for message regulation and relevant processing in the label distribution. An LSR can directly map the routing information on the network layer to a switch path on the data link over LDP, and then establish an LSP on the network layer. The LSP can be set up between two adjacent LSRs or terminated at an Egress LSR. All the LSRs in between adopt the label switching LDP Basic Concepts I. LDP Peers LDP peers refer to two LSRs undergo an LDP session by exchanging label/fec mapping information over LDP. The LDP peers can obtain the other s label information through an LDP session, namely, the LDP is bidirectional. 1-6

9 Chapter 1 MPLS Architecture II. LDP Session An LDP session is to exchange label and release messages between LSRs. There are two types of LDP session: Local LDP Session: an LDP session between two directly connecting LSRs. Remote LDP Session: an LDP session between two indirectly connecting LSRs. III. LDP Message There are four types of message involved in the LDP. Discovery message: used to notify or maintain the existing LSRs in the network; Session message: used to establish, maintain or terminate a session between LDP peers; Advertisement message: used to establish, modify or delete a flag, that is, an FEC binding; Notification message: used to provide suggestive messages or error notifications. IV. Label Space and LDP Identifier A label space refers to the range of labels that can be allocated to LDP peers. You can specify a label space for each interface of an LSR or for the entire LSR. An LDP identifier is to identify a special LSR label space. It is a six-byte value in the following format: <IP address>: <Label space number> Here the IP address of four bytes is the LSR IP address and the remaining two bytes is the label space number LDP Working Process Figure 1-5 illustrates the LDP label distribution. 1-7

10 Chapter 1 MPLS Architecture LSP1 Ingress A B C Egress LSP2 E Label Request Label Mapping F G D MPLS LSR MPLS LER H LDP Session Figure 1-5 Label distribution process On an LSP, along the data transmission direction, neighboring LSRs are respectively called as upstream LSR and downstream LSR. On LSP1 shown in Figure 1-5, LSR B is the upstream LSR of LSR C. Labels can be distributed in two modes: downstream on demand (DoD) and downstream unsolicited (DU), depending on whether label mapping distribution is done at the downstream with solicitation or without. 1) DoD mode In DoD mode, the label is distributed in this way: the upstream LSR sends label request message (containing FEC descriptive information) to the downstream LSR, and the downstream LSR distributes label for this FEC, and then sends the bound label back to the upstream LSR through label mapping message. When the downstream LSR feeds back the label mapping message depends on whether this LSR uses independent label control mode or sequential label control mode. When the sequential label control mode is used by the downstream LSR, the label mapping message is sent back to its upstream LSR if only it has received the label mapping message from its downstream LSR. And when the independent label control mode is used by the downstream LSR, it will send label mapping message to its upstream LSR immediately, no matter if it has received the returned label mapping message from its downstream LSR. Usually, the upstream LSR selects the downstream LSR according to the information in its routing table. In Figure 1-5, the sequential label control mode has been used by the LSRs on the way along LSP1, and the independent label control mode has been used by the LSRs on LSP2. 2) DU mode 1-8

11 Chapter 1 MPLS Architecture In DU mode, the label is distributed in the following way: when LDP session is established successfully, the downstream LSR will actively distribute label mapping message to its upstream LSR. The upstream LSR saves the label mapping information and processes the received label mapping information according to the routing table LDP Basic The basic LDP operation includes: Discovery phase Session establishment and maintenance LSP setup and maintenance Session termination I. Discovery Phase The originating LSR periodically sends a Hello message to its adjacent LSRs, notifying them its peer information, so that the LSR can automatically find its LDP peer. There are two types of LDP discovery mechanisms. Basic discovery mechanism The basic discovery mechanism is to discover the local LDP peer, that is, to establish a local LDP session between directly connecting LSRs. In this case, the LSR periodically sends a Hello message of the LDP link to a specific port, carrying the LDP identifier of the label space where the specific port belongs as well as other relevant information. If the LSR receives the Hello message over the specific port, it knows that there is a potential reachable peer on the link layer and learns the label space of the port. Extended discovery mechanism The extended discovery mechanism is to discover a remote LDP peer, that is, to establish a remote LDP session between non-directly connecting LSRs. In this case, the LSR periodically sends an LDP Targeted Hello message to a specific IP address. The LDP Targeted Hello message is sent in a UDP packet to the Well-known LDP discovery port of the specific address. The message contains the desired label space of the LSR as well as all relevant information. II. Session Establishment and Maintenance After the peer is set up, the LSR begins to establish a session by the following two steps: Establishing a connection on the transport layer, that is, establishing TCP connection between the LSR peers; 1-9

12 Chapter 1 MPLS Architecture Initializing the session and negotiating the parameters involved in the session, such as the LDP version, label distribution mode, timer timeout and label space. III. LSP Setup and Maintenance Actually, LSP establishment refers to the process of binding FEC with the label, and then advertising this binding to the adjacent LSR on LSP. This process is implemented through LDP in the following steps: 1) When the routing of the network changes and an LER finds a new destination address in its routing table not belonging to any existing FEC, the LER needs to create an FEC for the address and determine routes for the FEC, and then sends a label request message to the downstream LSR, indicating the FEC to be allocated; 2) After the downstream LSR receives the label request message and records it, it relays the message to the next hop LSR according to its routing table; 3) When the label request message reaches the destination LSR or the Egress LSR in the MPLS network and either can allocate the requested label, it will allocate the label to the FEC after the label request message passes its authentication. Then it sends a label mapping message to the upstream LSR with the allocated label information included; 4) The upstream LSR compares the received label mapping message with its label database, allocates the matched label to the FEC, adds the map to its label forwarding table, and then sends the label mapping message to its upstream LSR; 5) When the Ingress ISR receives the label mapping message, it adds the map to its label forwarding table. In this way, an LSP is set up and the corresponding FEC data packet can be forwarded based on its label. IV. Session Termination The LDP checks the session integrity depending on the LDP PDU transmitted in the session connection. The LSR sets up a living timer for each session and refreshes the timer after receiving an LDP PDU. If the timer expires before the reception of an LDP PDU, the LSR considers the session interrupted and tears down the corresponding connection on the transport layer to terminate the session LDP Loop Detection It is necessary to prevent path loop from happening while establishing LSP in the MPLS domain. The LSP loop detection mechanism can detect such path loop and avoid message loop occurring such as the label request message. To avoid the LSP path loop, two methods can be used: 1-10

13 Chapter 1 MPLS Architecture I. Maximum Hop Count The maximum hop count method is to contain the hop-count information in the message bound with the forwarding label. This value is added by one for each hop. When the value exceeds the specified maximum value, it is considered that a loop happens, and the process for establishing LSP is terminated. II. Path Vector The path vector method is to record the path information in the message bound with the forwarding label. For every hop, the corresponding router checks if its ID is contained in this record. If not, the router adds its ID into the record; and if so, it indicates that a loop happens and the process for establishing LSP is terminated Constrain-based Routing LDP MPLS also supports Constrain-based Routing LDP mechanism (CR-LDP). The so-called CR-LDP refers to that, while originating the establishment of LSP, the ingress node adds some constrain information for LSP routing in label request message. Two routing approaches are available: strict explicit routing, where the constraint information is the exact designation of all the LSRs along the path; and loose explicit routing where only some of the LSRs along the path are specified. 1.5 MPLS and Other Protocols MPLS and Routing Protocols When LDP establishes LSP in hop-by-hop mode, the next hop will be determined by using the information that is usually collected via such routing protocols as IGP, BGP in each LSR route forwarding table on the way. However, LDP just uses the routing information indirectly, rather than being associated with various routing protocols directly. On the other hand, although LDP is the special protocol for implementing label distribution, but it is not the sole protocol for label distribution. The existing protocols such as BGP, RSVP, after being extended, can also support MPLS label distribution. For some MPLS applications, it is also necessary to extend some routing protocols. For example, MPLS-based VPN application needs the extension of BGP so that the BGP is capable of supporting the sending of VPN routing information. In addition, MPLS-based Traffic Engineering (TE) needs the extension of OSPF or IS-IS protocol to carry link status information MPLS Extension by RSVP Resource Reservation Protocol (RSVP), after being extended, can support MPLS label distribution. At the same time, while transmitting label-binding message, it is also 1-11

14 Chapter 1 MPLS Architecture capable of carrying resource reservation information. The LSP established in this way is of resource reservation function, namely, the LSRs on the way can distribute some resources for this LSP to ensure the service transmitted on it. The extension of RSVP mainly refers to adding new objects in its Path message and Resv message. Besides carrying label binding information, these new objects are also capable of carrying the constrain information for searching path for the LSRs on the way, thus supporting LSP constraining function on routing. The extended RSVP also supports fast rerouting, namely, when it is necessary to change LSP under some condition, the original service flow can be rerouted to the newly established LSP without interrupting the customer service. 1.6 MPLS Application MPLS-Based VPN For traditional VPN, the transmission of the data flow between private networks on the public network is usually realized via such tunneling protocols as GRE, L2TP and PPTP, and LSP itself is the tunnel on the public network. The implementation of VPN using MPLS is of natural advantages. MPLS-based VPN connects the geographically different branches of private network by using LSP, forming a united network. MPLS-based VPN also supports the interconnection between different VPNs. PE3 CE3 Branch 3 of private network Branch 1 of private network CE1 PE1 Backbone network CE2 Branch 2 of private network PE2 Figure 1-6 MPLS-based VPN The basic structure of MPLS-based VPN is shown in Figure 1-6. CE is the customer edge device, and it may be a router, a switch, or perhaps a host. PE is a service provider edge router, which is located on the backbone network. PE is responsible for the management of VPN customers, establishing LSP connection between various PEs, route allocation among different branches of the same VPN customer. Usually the route allocation between PEs is realized by using extended BGP. MPLS-based VPN supports the IP address multiplexing between different branches and the interconnection between different VPNs. Compared with traditional route, it is necessary to add branch and VPN distinguisher information in VPN route. Therefore, it is necessary to extend BGP to carry VPN routing information. 1-12

15 Chapter 1 MPLS Architecture MPLS-Based Traffic Engineering I. Application of Traffic Engineering Network congestion is the main problem affecting the backbone network performance. Usually the network is congested due to insufficient network resources or partially due to unbalanced network resources. Traffic Engineering is used to solve the congestion due to unbalanced load. Through monitoring network traffic and the load on network element dynamically, then adjusting traffic management parameters and routing parameters as well as resource constraining parameters in real time, the network prevents the congestion due to unbalanced load and the network resources is optimized. II. Advantages of mpls-based traffic engineering The existing IGPs are all driven by topology, and only the static connection of the network is taken into account. However, such dynamic status as bandwidth and traffic characteristics cannot be reflected. This is just the main reason resulting in unbalanced network load. MPLS, which is different from those of IGP, just satisfies the requirement of traffic engineering: MPLS supports the explicit LSP routing that is different from routing protocol path. Compared with traditional single IP packet forwarding, LSP is more convenient for management and maintenance. Constrain-routing-based LDP is capable of realizing various policies of traffic engineering. In addition, the system overhead of MPLS-based traffic engineering is even lower than that of other implementation modes. III. MPLS-based traffic engineering implementation While realizing traffic engineering by using MPLS, first it is necessary to generate MPLS derived graph according to the topology of physical network, i.e., the derived topology chart composed of such three elements as LSR, the LSP connecting LSR, and LSP attribute. Meanwhile, the data passing through MPLS domain are divided into several traffic trunks. Usually the traffic trunks are defined as all one-way traffic passing through Ingress and Egress in MPLS domain. Traffic trunks are of multiple attributes, including traffic parameters, path selection and maintenance mode, precedence, capability of preemption, resource affinity and so on. Some other attributes are also defined for the resources, such as resource level, maximum distribution multiplexing degree and so on. Then, the traffic trunk attributes, resource attributes and network status information are used as the policies for generating constrain route to find out traffic trunk path. In addition, the traffic trunk path can be adjusted dynamically according to network status variation MPLS-Based QoS QoS is indispensable to the implementation of voice, video, and some other real-time services over an IP network in the sense that it can differentiate the data streams so 1-13

16 Chapter 1 MPLS Architecture that those crucial, sensitive, and delay-sensitive data streams over the network can be processed first. As H3C devices support MPLS-based Diff-serv features, they can provide differentiated services for the data streams assigned with different precedence levels while maintaining high network efficiency. Thus, they can provide the services featuring guaranteed bandwidth, low delay, and low loss rate for voice and video traffic. As it is difficult to deploy TE over the entire network, the Diff-ser model is always preferred for implementing QoS in actual networking solutions. The basic mechanism of Diff-serv goes like this: A service is mapped to a service category, which can be uniquely identified by the DS bits in the ToS field of the IP packets, according to the required service quality at the network edge. Then, the nodes on the backbone network adopt the proper service policy to process the packets according to the service category defined by the DS bits (derived from the ToS field), ensuring the proper service quality. The service quality classification and the label mechanism in Diff-serv are very similar to the label distribution in MPLS. In fact, MPLS-based Diff-serv is fulfilled by integrating DS assignment into the label distribution process of MPLS. Diff-serv defines the same processing method, which includes queue selection, queuing, and the drop operation, for each service category. The combination of these processing operations is called Per Hop Behavior (PHB). In addition, the packets that belong to the same PHB may be assigned with different drop preference. The information of PHB and drop preference is indicated by the DS code assigned to the packets. The DS codes are also known as Diff-serv Code Point (DSCP). For more information about Diff-serv, refer to the Section QoS module in this manual. The following methods are available for supporting end-to-end QoS based on the Diff-serv model: IP Precedence for Traffic Classification IP precedence classification is implemented at the network edge. It makes use of the 3-bit Type-of-Service field in the IPv4 header to sort precedence of the IP packets according to their addresses. At the core, different queuing technologies are used to make different processing on the streams of different precedence, thus to discriminate the services at different levels. To implement Diff-serv for the voice, image, and data streams, when PE labels the packets, that is, makes label switching for different traffic, it will map the ToS value carried in the IP packets to the CoS field of the label. Thus, the type information previously carried by IP will be carried by the label. Depending on the CoS field of the label, PE routers will implement differentiated scheduling on the packets using PQ, CQ, WFQ or CBQ. TP for implementing committed and constraint bandwidth This function can be implemented by configuring Traffic Policing (TP) on the link that connects PE to CE. In addition, TP also provides the functions of committed bandwidth and constraint bandwidth. 1-14

17 Chapter 1 MPLS Architecture WRED for congestion avoidance WRED can be used to monitor and alleviate network congestion at the bottleneck of the network. Usually, congestion is more likely caused at the access layer. At the onset of congestion, WRED, which is monitoring the network load, begins to select packets and discard them for decreasing the traffic size. The packet drop policy of WRED is to discard the packets of low preference to ensure the smooth transmission of high preference packets. Running WRED on a port prone to congestion is a good choice for congestion avoidance. In actual deployment, the tasks should be distributed in order to achieve the optimal efficiency. As QoS is such an application that will consume enormous resources of the processor, the tasks of QoS are shared by the edge and core routers so as to alleviate the load imposed on a single router. To sum up, four steps should be followed in order to implement CoS-based Diff-serv: Impose the incoming bandwidth constraint on the MPLS edge router to classify the incoming traffic. Adopt CAR on the edge devices so that they can share the work of bandwidth management. The MPLS core router accomplishes CoS management to implement Diff-serv QoS. The egress device, like the ingress device, implements bandwidth restriction. The bandwidth restriction implemented by the ingress and egress devices protects the network from congestion and hence significantly improves the network scalability. For more details, see QoS module. 1-15

18 2.1 Introduction to MPLS Configuration The following MPLS basic capabilities are available: Basic MPLS forwarding Comware supports the basic MPLS forwarding function, including label packet forwarding and TTL processing. LDP session establishment and LSP maintenance Each interface supports LDP session and supports the loop detection in both maximum hop count mode and path vector mode. It can create and delete LSPs. It also supports loose routing and strict explicit routing: You can also specify LSPs. Besides MPLS basic functions, Comware also provide performance monitoring and fault diagnostic tools. To enable basic MPLS functions at a firewall, you should complete the following configuration tasks: 1) Configure LSR ID 2) Enable MPLS 3) Enable LDP protocol 4) Enter interface view and enable interface LDP function Then the firewall can provides MPLS forwarding and LDP signaling functions. If you want to modify the default parameters or enable some special functions, for example, creating LSP, creating explicit route, you just configure according to the method in configuration list. For some complicated functions, configuration combination may be required. 2.2 MPLS Basic Capability Configuration MPLS basic capability configuration (compulsory) includes: Define MPLS LSR ID Disable/enable LDP and enter LDP view MPLS basic capability configuration (optional) includes: Enable LDP on interface Control LDP loop detection Set LDP session keepalive parameters on interface 2-1

19 2.2.1 Defining MPLS LSR ID Before configuring any other MPLS command, it is necessary to configure LSR ID firstly. The ID is usually in IP address format and must be unique in the domain. Perform the following configuration in system view. Table 2-1 Define MPLS LSR ID Define LSR ID Delete LSR ID mpls lsr-id ip-address undo mpls lsr-id By default, LSR ID is not specified Entering MPLS View Execute the mpls command in system view to enter the MPLS view. Then, you can perform the MPLS configurations. Table 2-2 Enter MPLS view Enter MPLS view Disable MPLS globally. mpls undo mpls Executing the mpls command in interface view will enable the MPLS capability on the corresponding interface. For those link layer protocols that do not support broadcast, such as X.25, Frame Relay, ATM, you must use the protocol ip { ip-address [ ip-mask ] default inarp [ minutes ] } [ broadcast ] command to configure the broadcast attribute to support the transmission of broadcast and multicast packets Configuring the Topology-Driven LSP Setup Policy It refers to specifying filtering policy as all or ip-prefix. Perform the following configuration in MPLS view. Table 2-3 Configure a topology-driven LSP setup policy Configure a topology-driven LSP setup policy lsp-trigger { all ip-prefix ip-prefix } 2-2

20 Disable the filtering condition specified by the arguments and no routes of any type can trigger the LSP setup undo lsp-trigger { all ip-prefix [ ip-prefix ] } Configuring Static LSP You can manually set an LSR to be a node along an LSP, and place a limit on the traffic over the LSP. Depending on the position in an MPLS domain, LSR can be ingress, transit node, or an egress. Note that the correct operation of this LSP can be ensured only after the LSRs along the specified LSP have been properly configured. The undo static-lsp command is used to delete a specified LSP established manually. Perform the following configuration in MPLS view. Table 2-4 Set this LSR to be a node on a specified LSP Set the current LSR to be the ingress of a specified LSP Set the current LSR to be a transit node along the specified LSP Set the current LSR to be the egress of the specified LSP static-lsp ingress lsp-name destination dest-addr { addr-mask mask-length } } { nexthop next-hop-addr outgoing-interface interface-type interface-num } out-label out-label-value undo static-lsp ingress lsp-name static-lsp transit lsp-name incoming-interface interface-type interface-num } in-label in-label-value { nexthop next-hop-addr outgoing-interface interface-type interface-num } out-label out-label-value undo static-lsp transit lsp-name static-lsp egress lsp-name incoming-interface interface-type interface-num in-label in-label-value undo static-lsp egress lsp-name Configuring IP TTL Duplication of MPLS The MPLS label contains an eight-bit TTL field same as the one used in an IP header. It can be used for supporting the tracert command in addition to suppressing routing loops. As described in RFC3031, when an LSR labels a packet, it needs to copy the TTL value in the IP packet or the upper layer label into the TTL field in the added label. When it forwards a labeled packet, it decrements the TTL value in the top label by one. When the LSR pops the stack, it copies the TTL value in the top label back to the IP packet or the lower label. 2-3

21 If an LSP has a non-ttl LSP segment which comprises a sequence of ATM-LSRs or FR-LSRs unable to handle the TTL field, the TTL value must be decremented by the value that reflects the length of the non-ttl LSP segment before a packet is forwarded into this segment. In MPLS VPN networking, you may hide the MPLS backbone topology for security sake. You cannot however, apply on the ingress TTL duplication to VPN packets. Perform the following configuration in MPLS view. Table 2-5 Configure IP TTL duplication of MPLS Enable IP TTL duplication of MPLS. ttl propagate { public vpn } Disable IP TTL duplication of MPLS. undo ttl propagate { public vpn } By default, IP TTL duplication is enabled for public-network packets and disabled for VPN packets. If IP TTL duplication is enabled at the ingress, the TTL value of packets is decremented at each LSR hop that it passes. This allows the tracert to reflect the path that a packet travels. If IP TTL duplication is disabled at the ingress, the TTL value is not decremented at the LSR hops that packets travel and as such, those hops in the MPLS backbone are excluded from the path shown after you run a tracert, just as if the ingress and the egress are directly connected. Note that: Inside an MPLS domain, if an MPLS packet has a label stack, the TTL value in a label is always copied from other labels in the stack. The TTL value of the transmitted local packets is copied regardless whether IP TTL duplication is enabled or not. This ensures that the local administrator can execute the tracert command to test the network. At the egress, if IP TTL duplication is enabled, the TTL value in the MPLS label is copied to the TTL field in the IP header and decremented by one. Configure IP duplication for VPN packets in the same way at the ingress and the egress: if the ttl propagate vpn command is enabled at the ingress, enable it at the egress; if the command is disabled at the ingress, disable it also at the egress. This ensures that the results of traceroutes can reflect the real network conditions. You are recommended to enable this function on the involved PEs to ensure consistency of the results gotten by tracerting on different PEs. You need not to configure the ttl propagate command on any P routers. 2-4

22 2.2.6 Configuring MPLS to Return ICMP Responses by IP routing In an MPLS VPN network, a P router cannot route the IP packets encapsulated in MPLS. When the TTL value of an MPLS packet expires, the ICMP response continues to travel the LSP until reaching the egress, where it is forwarded by IP routing. This approach increases network traffic while decreasing reliability of packet forwarding. For a one-tier MPLS packet with TTL expired, you can configure to have its ICMP response forwarded by local IP routing, that is, the default. Perform the following configuration in MPLS view. Table 2-6 Configure MPLS to return ICMP responses by IP routing Return ICMP responses by IP routing. Return ICMP responses along the LSP. ttl expiration pop undo ttl expiration pop On an ASBR or SPE (it can be an SPE in a nesting application) on an HoVPN network, the MPLS packets that carry VPN packets may have only one-tier labels. In this circumstance, to tracert to a VPN to view the forwarding path of the routers on the public network, you need to perform the following tasks: 1) Configure the ttl propagate vpn command on all the involved PEs. 2) Configure the undo ttl expiration pop command on the ASBR and SPE to guarantee ICMP responses to be forwarded along the original LSP. 2.3 LDP Configuration Perform the following compulsory tasks in configuring LDP: Enable LDP Enable LDP on Interface Perform the following optional tasks in configuring LDP: Configure LDP extended discovery mode Configure the label of the penultimate hop at the egress Configure session parameters Configure LDP loop detection control Configuring LDP authentication mode Enabling/Disabling LDP To configure LDP, first enable LDP. Perform the following configuration in system view. 2-5

23 Table 2-7 Enable/disable LDP view Enable LDP Disable LDP mpls ldp undo mpls ldp By default, LDP is disabled Enabling/Disabling LDP on Interface To make the interface be of MPLS function, it is necessary to enable the LDP of interface in the interface view. Then the interface is capable of establishing LDP session and forwarding labeled packet. Disabling LDP function will delete LDP peer, delete LSP tunnel, close TCP connection, interrupt peer LDP session, stop sending HELLO packets, and stop labeled packet forwarding. Perform the following configuration in interface view. Table 2-8 Enable/disable LDP on interface Enable LDP function on interface Disable LDP function on interface mpls ldp enable mpls ldp disable By default, the interface MPLS function is disabled Configure LDP extended discovery mode It is to create extended discovery mode, to set up sessions with peers not directly connected to the link. I. Entering remote-peer mode Perform the following configuration in system view. Table 2-9 Enter the extended discovery mode Enter the extended discovery mode Delete the corresponding remote-peer mpls ldp remote-peer index undo mpls ldp remote-peer index There is no default remote-peer. 2-6

24 II. Configuring a remote-peer address You can specify the address of any LDP-enabled interface on the remote-peer or the address of the loopback interface on the LSR that has advertised the route as the address of the remote-peer. Perform the following configuration in remote-peer view. Table 2-10 Configure a remote-peer address Configure the remote-peer address remote-ip ip-address There is no default remote-peer Configuring Session Parameters I. Configuring session hold-time The LDP entity on the interface sends Hello packets periodically to find out LDP peer, and the established sessions must also maintain their existence by periodic message (if there is no LDP message, then Keepalive message must be sent). Caution: Modifying the holdtime argument does not bring changes for the value of the original session. However, the modification will take effect on sessions that are going to be established. Here the session refers to link session, but not remote session. Perform the following configuration in interface view. Table 2-11 Set interface LDP session parameters Set interface LDP session keepalive parameters Restore the default interface LDP session parameters mpls ldp timer { session-hold session-holdtime hello hello-holdtime } undo mpls ldp timer { session-hold hello } The session-holdtime argument defaults to 60 seconds and the hello-holdtime argument defaults to 15 seconds. 2-7

25 II. Configuring hello transport-address The transport-address discussed here refers to the address carried in the transport address field TLV in hello messages. Generally, transport-address is the MPLS LSR ID of the current LSR, but other configurations may be required in some applications. Perform the following configuration in interface view. Table 2-12 Configure hello transport-address Configure a hello transport-address Restore the default hello transport-address mpls ldp transport-ip { interface ip-address } undo mpls ldp transport-ip Transport-address defaults to the MPLS LSR ID of the current LSR. When MPLS LDP is enabled on multiple directly connected links, these links must be configured with the same transport address. It is recommended that you take the default LSR-ID as the transport address; otherwise, the LDP session may not be well established Configuring Remote LDP Peer Session Parameters I. Configuring hello timers Perform the following configuration in remote-peer view. Table 2-13 Configure hello timers Configure the interval and hold-time value for hello packets Restore the interval and hold-time value for hello packets to the default mpls ldp timer targeted-hello { interval time holdtime time } undo mpls ldp timer targeted-hello { interval holdtime } By default, the hello interval is 13 seconds; the hold-time is 45 seconds. II. Configuring session keepalive timers Perform the following configuration in remote-peer view. Table 2-14 Configure session keepalive timers Configure the interval and hold-time value for session keepalive packets mpls ldp timer targeted-session-hold { interval time holdtime time } 2-8

26 Restore the interval and hold-time value for session keepalive packets to the default undo mpls ldp timer targeted-session-hold { interval holdtime } By default, the keepalive interval is 24 seconds; the hold-time is 60 seconds LDP Loop Detection Control I. Enabling loop detection It is used to enable or disable the loop detection function during LDP signaling process. The loop detection includes maximum hop count mode and path vector mode. In maximum hop count mode, hop count information is included in label binding information and is added by 1 when the packet passes through another hop. When the value exceeds the maximum, it is reckoned that loop appears, so LSP setup process terminates. In path vector mode, path information is included in the label binding information. Each time the packet passes through a hop, the corresponding router will check whether its ID is recorded in the path information. If not, the router just adds its ID. If yes, it means loop appears, so LSP setup process terminates. Perform the following configuration in system view. Table 2-15 Enable loop detection Enable loop detection Disable loop detection mpls ldp loop-detect undo mpls ldp loop-detect By default, the loop detection is disabled. II. Setting the maximum hop count for loop detection When maximum hop count mode is adopted for loop detection, the maximum hop-count value can be defined. And if the maximum value is reached, it is considered that a loop happens and the LSP establishment fails. Perform the following configuration in system view. 2-9

27 Table 2-16 Set the maximum hop count for loop detection Set maximum hop count for loop detection Restore default maximum hop count mpls ldp hops-count hop-number undo mpls ldp hops-count The maximum hop count defaults to 32. III. Setting the maximum value of the path vector When path vector mode is adopted for loop detection, it is also necessary to specify the maximum value of LSP path. In this way, when one of the following conditions is met, it is considered that a loop happens and the LSP establishment fails. 1) The record of this LSR already exists in the path vector recording table. 2) The path hop count reaches the maximum value set here. Perform the following configuration in system view. Table 2-17 Set the maximum value of the path vector Set maximum value of the path vector Remove maximum value setting for path vector mpls ldp path-vectors pv-number undo mpls ldp path-vectors The maximum hop count defaults to Configuring LDP Authentication Mode Perform the following configuration in interface view or remote-peer view. Table 2-18 Configure LDP authentication mode Configure LDP authentication mode Remove LDP authentication mpls ldp password { cipher simple } password undo mpls ldp password 2-10

28 2.4 Displaying and Debugging MPLS Displaying and Debugging MPLS MPLS provides abundant display and debugging commands for monitoring LDP session state, tunnel, all the LSPs and their states, and so on. These commands are the powerful debugging and diagnosing tools. I. Displaying static LSPs After accomplishing the configuration tasks mentioned earlier, you can execute the display command in any view to view the running state of a single or all the static LSPs and thus to evaluate the effect of the configurations. Table 2-19 Display the static LSP information Display the static LSP information display mpls static-lsp [ verbose ] [ include text ] II. Displaying MPLS statistics Execute the display command in any view or the reset command in user view to view or clear statistics about the specified or all static LSPs. Table 2-20 Display/clear the MPLS statistics Display the MPLS statistics Clear the MPLS statistics display mpls statistics { interface { all interface-type interface-number } } { lsp [ lsp-index all name lsp-name ] } } reset mpls statistics { { interface { all interface-type interface-num } } { lsp lsp-name } } III. Displaying MPLS fast-forwarding information Execute the display command in any view and reset command in user view to view or clear MPLS fast-forwarding information. Table 2-21 Display/clear MPLS fast-forwarding information Display MPLS fast-forwarding information Clear MPLS fast-forwarding information display mpls fast-forwarding cache reset mpls fast-forwarding cache 2-11