Multiprotocol Label Switching (MPLS) Migration

Network Management

Every global organization has to handle network complexities with their increasing number of remote sites. This is where network performance factors such as connectivity, data transfer, and traffic management gain more significance. While several types of traffic flow within the network (such as multimedia, email messages, VoIP, SAP), enterprises were pushed to use private lines for voice, Frame Relay and ATM for data access. Each type of data has increased exponentially and the dependency on these datalink protocols has let companies to explore on a solution that serves as a common platform to manage all types of traffic. With requirements of bandwidth scalability, security, flexibility, and fast data transmission, MPLS (Multi Protocol Label Switching) outdoes others as a technique that eliminates the discrepancies found in Layer-2 protocols.

Why do companies migrate towards MPLS?

The following key factors determine why enterprises migrate towards this technology:

• Scalability
• Security
• Cost
• Speed
• Quality
• Reliability

Scalability: This feature defines why MPLS is most sought by companies when they plan to migrate. Common protocols such as Frame Relay and ATM work with a specified bandwidth for T3 and T1 connections (45Mbps/1.5Mbps). In case of MPLS, there are no specifications for bandwidth as this label switching technology can scale from a very low range to the highest bandwidth as the network expands. The specialty is that it purges cell overhead of ATM, segmentation, and reassembling of conventional protocols. It allows optimum bandwidth usage and is flexible with T1, T3, satellite, and wireless connections. In addition, distribution protocols such as LDP, RSVP augment packet relay with routing protocols like OSPF, IS-IS for automatic rerouting. The protocol support of IP versions (IPv4, IPv6), ATM, Frame Relay, FDDI, and Ethernet, extends its adaptability. Huge networks can be managed with lesser number of labels for packet forwarding, such that they scale to corporate firm’s network expansion.

Security: As companies expand with several local and remote sites globally, MPLS gains high magnitude as security has to be ensured in both edge and core of the network. It does this by concealing core network information from customer VPNs to avoid hacking of internal information. Secure packet transmission is possible only when it facilitates protection mechanisms such as validation of packet routing, label signaling, encryption, and decryption. Using virtual private circuits, this technology removes the need of data encryption as seen in IP based networks. In each of these, this label-based technique provides tight security through generic routing encapsulation (GRE), segregating Virtual Private Network (VPN) and Internet traffic, packet filtering, defend from VPN spoofing, setting firewall, and gateway to prevent denial-of-service attacks. The infrastructure is set in such a way that the MPLS core is independent from other VPNs and hides them external network access.

Implementation Cost: Most of the modernized firms use T1 and T3 connections that are quite expensive for their complex network operations. For additional bandwidth requirements, extra purchase of these tiered connections incurs more investment for any corporate firm during the capacity-planning phase. This increases the cost of ownership (TCO) which eventually trims down the ROI return. A smart way to deal with this is MPLS, a ‘best managed service’ unifying all networks on a single platform with its uncomplicated deployment. It is also protocol-neutral and renders full support for applications ranging from business critical traffic (VoIP, SAP) to mail attachments. With less resources and maintenance, this assures application delivery even at traffic congestion. Rather than investing over private lines (on lease), Frame Relay, and ATM, MPLS requires only a one-time investment to merge all protocols with its core infrastructure, which simplifies the complexity. By sending VoIP and video traffic over data and control planes, it shrinks the operating cost of the management and further purchase made by the management on data services is avoided. For the reason that it is implemented worldwide, its cost is experiencing a lowdown. Hence, it is evident that enterprises planning to migrate will reap the benefits of low TCO and high ROI for their full mesh network topology as they expand.

Packet Transfer Rate: Using labels for packet routing, MPLS allows high-speed packet transfer with its ‘label-switching’ technique. Companies that use Internet Protocol (IP) in their network undergo a conventional packet forwarding mechanism. For each hop between nodes, packets will be analyzed, read header information, and finally lookup in routing table takes place. This is a time-eating process as each hop is dependent on a constant lookups in route tables. MPLS is a total contrast to this where packets are assigned with labels on each node and headers are analyzed only once for forwarding packet through label switch paths (LSP). As it is ‘label based’, packet relay is quick that routes them in the configured path for fast packet transmission With this mechanism, packet that requires zero-delay transmission can make it even at traffic blockage, as a result eradicating the occurrence of packet dropping of sensitive traffic. Forward Equivalence Class (FEC) determines the type of traffic and subsequently a label is assigned according to the next label switch router (LSR). Thus, performance metrics such as processing speed, response time, and packet transfer rate are comfortably met.

Class of Service (CoS): This feature is used to handle traffic in both virtual and point-to-point circuits. Both Frame Relay and ATM protocols use ‘service class’ to set type of traffic for packets in point-to-point circuits. Such circuits requires heavy engineering tasks to provision the entire network. Most of the IP implemented networks are not scalable and for this reason, a managed service like MPLS meets the requirement of companies with more of real-time VoIP and SAP traffic. A CoS value (3-bit) is encoded in the header to route traffic based on LSP, after placing them in the priority queue. Managing networks based on CoS makes it simple with easy-to-do engineering tasks for network. It also comprises methods to reserve resources for specific business applications by dropping delayed packets, thus making it available for data transmission. At this instant, the traffic carrying multimedia content (audio/video - have a huge impact over packet delivery) gains the most from MPLS as it reduces latency factors such as RTT and jitter.

Reliability& Redundancy: The constraints of transmission such as bandwidth allocation, resource usage, and packet relay determine the reliability of any protocol. Even at times of primary link failures, Multiprotocol Label Switching (MPLS) manages these bottlenecks with automatic rerouting of packets in alternative path of backup tunnels and utilizes bandwidth to the best possible level showing 99.999% reliability. It makes certain that backup premises are able to reconnect with remote sites during disasters. This ensures disaster recovery and any overflowing traffic in a link should be efficiently handled by the traffic engineering tasks set for a protocol. Any network operation should be done with no delay, as any failures will be handled with an alternate setup. Most of the conventional protocols do not conform to these standards, which makes MPLS the finest option to overcome all types of traffic bottlenecks.

Request for a free white paper on “Pitfalls to Avoid while Migrating to Multiprotocol Label Switching (MPLS)”.

 

---