IP-MPLS: A Routing Mechanism

CHAPTER VIII IP-MPLS ==================== 0. **Introduction:** 1. Multiprotocol label switching (MPLS) also known as IP-MPLS (IP- MPLS is one type of MPLS which primarily uses IP-routers) is a routing mechanism within a telecommunications network. The routers direct data from one nod...

CHAPTER VIII IP-MPLS ==================== 0. **Introduction:** 1. Multiprotocol label switching (MPLS) also known as IP-MPLS (IP- MPLS is one type of MPLS which primarily uses IP-routers) is a routing mechanism within a telecommunications network. The routers direct data from one node to another based on the short path labels instead of the relatively longer network addresses. It avoids the need for complex lookups in the routing table, so communications tend to be faster. Since it also encapsulates the protocols of the individual streams which are diverse, into the packets with its own protocol, it is called the "multiprotocol" routing technique. 2. MPLS is an "Internet Engineering Task Force" (IETF) specified framework that provides efficient forwarding, routing and switching of traffic flow through the network. 3. MPLS belongs to the family of packet switching networks and was designed to overcome the limitations of IP based forwarding for VPN. In a traditional IP network each router performs an IP lookup, determines the next hop based on its routing table and forwards the packet to the next hop thereby creating a lot of overhead at each routers interface. However, MPLS on the other hand makes packet forwarding decisions which are based entirely on the label of the packet without the need to examine the packet itself. 4. MPLS works in between OSI data link layer and network layer and is summarised as Layer 2.5 networking protocol. MPLS is an innovative approach that uses a label based forwarding model. 5. Out of the three major technologies viz. IP-MPLS, MPLS-TP and Carrier Ethernet, IP-MPLS has been chosen as the choice of future transport technology for Indian Railways. The basic advantage of IP- MPLS is its support for L2 and L3 services that is essential for Railways. Besides this, it also services the requirement of core, aggregation and access network and can work on a common NMS for OAM(Operations, Administration and Management/Maintenance) of all the three parts of a Network. The advantage of IP-MPLS is that it supports IP routing as well as network oriented connections. The forwarding is done through hardware with the introduction of MPLS and hence is much faster than normal routing. The paths are unidirectional and the forward and return paths are usually different. For serving the requirement of Transport, congruent bidirectional paths can be defined. SD-WAN ([software-defined networking](https://en.wikipedia.org/wiki/Software-defined_networking) in a [wide area network](https://en.wikipedia.org/wiki/Wide_area_network) ) with IP-MPLS simplifies the management and operation of a WAN by decoupling the networking hardware from its control mechanism. 6. Technology & Terminology: ========================= 1. MPLS is not associated with any specific technology, rather it is an overlay technique that aims to improve performance and efficiency. 2. When a packet enters the MPLS network, a forwarding equivalence class (FEC) value is assigned to it by adding a small label to the packet. Every router within the network knows how to handle different FEC labels, so there is no need to do a header analysis each time. Instead, every router uses the label as an index to identify a new FEC for that packet. 3. MPLS creates a predetermined path to route traffic in the most efficient way possible based on the FEC label. 4. This mechanism gives routers the option to choose low-latency routes for certain applications like live video streaming so it is delivered faster to the destination when compared to the traditional routing mechanism. 5. **Label:** The label is a part of MPLS header. It is placed between the data-link and IP headers. It identifies the path a packet should traverse. The MPLS header is composed of 32 bits out of which 20 bits are allocated to the label also called label stack, 3-bits are experimental bits often used for specifying class of service. One bit is reserved for the bottom of the stack bit and is set if no label follows. 8-bits are used for time-to-live (TTL) used in the same way like IP headers. 6. **Label Forwarding Information Base:** A table is created by a label switch-capable device that indicates where and how to forward frames with specific label values. 7. **Label Switched Path (LSP):** It is a unidirectional tunnel between a pair of routers routed across MPLS network. 8. **Label Edge Router/Ingress router (LER):** It is a router that first encapsulates the packet inside an MPLS LSP and also makes initial path selection. 9. **Label Switched Router (LSR):** A router which only does MPLS switching in the middle of an LSP. 10. **Egress Router:** The final router at the end of LSP which removes the label. 11. **Label switched:** When an LSR makes forwarding decisions based upon the presence of a label in the frame. 12. **Label Switch Controller (LSC):** An LSR that communicates with an ATM switch to provide and provision label information within the switch. 13. **Label Distribution Protocol (LDP):** It is one of the primary signalling protocols for distributing labels in MPLS network. It is a set of procedures and messages by which Label Switched Routers (LSR) establish Label Switched Path (LSP) through a network by mapping network layer routing information directly to data link layer switched paths. By Label Distribution Protocol, LSR can collect, distribute and release label binding information to other LSRs in the MPLS network thus enabling hop-by-hop delivery of packets in the network along routed paths. 14. **Forwarding Equivalence Class (FEC):** It is a group of IP packets that are forwarded. Packets within an FEC are equivalent in terms of forwarding, such as same destination, same path and same class of service. A LSP is assigned to each FEC that is defined using IP interior routing protocols.  7. MPLS Operation: =============== 15. MPLS relies on two principal components i.e. Control Plane and Data Plane. 1. **Control Plane:** Essential to MPLS is the notion of binding between a label and network layer routes. The control plane is responsible for the routing information exchange and label distribution between adjacent devices. It uses standard dynamic routing protocols, such OSPF routing, IS-IS and BGP, to exchange information with other routers to build IP forwarding table or label forwarding information base. The control component creates label bindings and then distributes the label- binding information among LSRs using a Label Distribution Protocol (LDP). 2. **Data Plane:** The data plane is responsible for forwarding packets according to the destination IP address or label using Label Forward Information Base (LFIB) managed by the control plane. The Data plane is a simple label based forwarding engine i.e independent of the type of routing protocol or label distribution protocol running on the control plane. 16. The network automatically builds routing tables as MPLS capable routers participate in interior gateway protocols (OSPF, IS-IS) throughout the network. 17. Label distribution protocol (LDP) establishes label to destination network mappings. Label distribution protocol (LDP) uses the routing topology in the tables to establish label values between the adjacent devices. This operation creates Label Switching Paths (LSP) pre- configured maps between destination end points. 18. A packet enters the ingress edge label switching router (LSR) where it is processed to determine which layer-3 service it requires, such as quality of service (QoS) and bandwidth management. The edge LSR selects and applies a label to the packet header and forwards it. 19. The LSR reads the label on each packet, replaces it with a new one as listed in the table and forwards the packet. 20. The Egress Edge Router strips the label, reads the packet header and forwards it to its final destination. 8. MPLS Services: ============== 21. **MPLS Traffic Engineering (MPLS-TE):** Traffic Engineering is the process of routing data traffic in order to balance the traffic load on various links, routers and switches in the network. 3. It has the ability to control specific routes across a network to reduce congestion and improves the cost of efficiency of carrying IP Traffic. MPLS is capable of full traffic engineering. 4. In MPLS TE, a Label Switched Path (LSP) is established for carrying traffic along an explicit traffic-engineered path, which can be different from the normal destination-based routing path. IP networks typically have multiple pathways that traffic can take to reach its destination. Relying solely on routing protocols such as Open Shortest Path First (OSPF) some of the paths may become congested while others are under-utilized. 5. MPLS can specify an explicit route for certain traffic flows such as Voice over IP (VoIP) to take less optimal but less congested routes and avoid packet loss while maintaining very high link utilization. 22. **MPLS and Quality of Service (QoS):** Some types of traffic, such as video, place specific demands on a network for successful transmission. QoS in an IP network gives devices the intelligence to preferentially handle traffic as dictated by each subscriber's network policy. 6. QoS is defined as those mechanisms that give network managers the ability to control the mix of bandwidth, delay, jitter and packet loss in the network. 7. At the ingress of the MPLS network, Internet Protocol (IP) precedence information can be copied as Class of Service (CoS) bits or can be mapped to set the appropriate MPLS CoS value in the MPLS label. 8. This is the distinction between IP QoS that is based on IP precedence field in the IP header and MPLS QoS that is based on the CoS bits in the MPLS label. 9. MPLS CoS information is used to provide differentiated services and MPLS CoS enables end-to-end IP QoS across the network. 23. **MPLS VPNs:** Virtual Private Networks (VPNs) are a method of interconnecting multiple sites belonging to a customer using a Service Provider (SP) backbone network in place of dedicated leased lines. 10. Each customer site is directly connected to the SP backbone. The SP can offer a VPN service more economically than dedicated private WANs built by each individual customer because the SP can share the same backbone network resources (bandwidth, redundant links) between many customers. 11. The customer also gains by outsourcing the complex task of planning; provisioning and managing a geographically distributed network to the SP. 12. MPLS-enabled IP VPNs are connectionless IP networks with the same privacy as frame relay and multiple IP service classes to enforce business-based policies. 13. MPLS-based VPNs make operations much more efficientthan the traditional overlay VPN solutions which requires tunnelling or encryption deployed over a frame relay, ATM or IP network. This mesh solution is built point-to-point, requiring separate configuration of each tunnel or Virtual Circuit (VC). Moreover, since traffic is tunnelled or overlaid, the circuit does not know which kind of traffic it carries. By contrast if the customer traffic can be classified by application type, such as voice, mission-critical applications or e-mail, the network can easily assign traffic to the appropriate VPN, without configuring complex, point-to-point meshes. 14. Compared to a VPN overlay solution, an MPLS-enabled VPN network can separate traffic and provide privacy without tunnelling or encryption. Using labels, MPLS enabled networks provide privacy on a network-by-network basis much as frame relay provides it on a connection-by-connection basis. 15. The frame relay VPN offers transport while MPLS-enabled network supports services. MPLS is the technology that brings - VPN awareness to switched or routed networks. 16. It enables quick and cost-effective deployment of VPNs of all sizes - over the same infrastructure. MPLS provides a flexible and elegant VPN solution based on the use of LSP tunnels to encapsulate VPN data. 9. Advantages and disadvantages: ============================= 24. **Advantages:** 17. **Scalable:** MPLS provides a highly scalable mechanism. It ensures high-performance telecommunication networks. Networks can easily be engineered and maintained for bandwidth optimization. 18. **Inter-Connectivity Growth:** MPLS allows growth of the inter- connectivity of the network by using the minimal addition of hardware. 19. **Common applications:** MPLS can be used for interconnecting data centers with branch offices and branches at other locations. 20. **Remote connections:** MPLS allows adding new remote connections without using any additional hardware system at the primary site. Being fully cloud-based, it doesn't require the point to point connectivity. 21. **WAN routing:** With the MPLS link, WAN routing is left to the service provider and employs fewer staff for WAN. 22. **Quality of service:** MPLS comes with the quality of service (QoS) options, which empowers to treat latency-sensitive traffic like VoIP etc. 23. **WAN protocol:** MPLS is the perfect mode to manage any-to-any connectivity, including video and voice. 24. **Service level agreements:** MPLS services are deliverable SLAs (service-level agreements). These SLAs include delivery guarantees unlike consumer broadband, etc. 25. **Enhanced bandwidth:** The technology allows accessing multiple traffic types. 26. **Improved up-time:** MPLS allows having an alternative network, thus improves up-time. 27. **Lower congestion:** With MPLS, there is an option to use alternative paths and avoid high traffic congestion, thus reduced network congestion. 25. Disadvantages: ============== 28. **Lack of Total Control:** The service provider has to configure the overall networks. And we will need to work along with service provider in routing MPLS traffic while using dynamic routing. MPLS does not allow having total control of the network. 29. **Expensive:** Since MPLS is an advanced way of networking, it can cost more than the Ethernet. However, the cost is less than T1 lines. 10. Implementation considerations: ============================== 26. From the various communications in use on the Indian Railways, a Division is the basic operational unit of the Railways and all the activities of all departments are initiated, implemented, coordinated and monitored and hence is the basic aggregation layer for the communication bandwidth. 27. Most circuits originate from Divisional HQ and terminate at one/many/all of the stations within the Division, adjacent divisional HQ, Zonal HQ and the internet gateway. In the event of any emergency or 28. Some of the existing data communication circuits are centralized across the zone such as UTS/FOIS. Some of the services/applications like IP exchanges, section control/TPC/TLC, applications on VOIP, File services for storage of critical data/drawings of departments, application software such as MS Office/AutoCAD/Primavera/Anti-Virus, VMS and Video Analytics for Surveillance etc.are controlled within the Division. 29. Considering the various services and applications used by the division, it is desirable that servers for running the various services and applications relevant to the Division are located in the Divisional HQs in suitable Data Centers. This will also serve to address latency and response time issues besides optimizing bandwidth utilization. 30. The specific implementation steps: 30. All future replacement of SDH shall only be with MPLS equipment. 31. Equipment with modular and hybrid interfaces should be used so that interfaces with legacy TDM equipment are replaced as and when needed. 32. As this is a new technology area, intensive training on these technologies should be imparted to officers and supervisors. Railways may also try to have dedicated staff and officers in the division for smooth adoption of these technologies. 33. Divisional and Zonal NOC for OAM of the unified network should be created. 34. NMS & NOC Architecture 1. NMS: ==== NOC: ==== 1. Zone HQ NOC capabilities: - Single point of contact for the interdivisional and inter-zonal issues. - Node installations, troubleshooting and updating for zonal nodes. - Service provisioning for zonal nodes. - Internet Policy Control. - Overall Performance reporting and improvement recommendations. - Patch management and whitelisting. - Backup management. 2. DIV HQ NOC capabilities: - Troubleshooting and updating. - Field support. - Node installations, troubleshooting and updating. - Service provisioning. - Performance reporting and improvement recommendations. - Patch management and whitelisting. - Backup management. -x-x-x- =======