Multi-Protocol Label Switching (MPLS) Conform…

Figure 5. Hyper-aggregation in conventional IP networks.

This is known as Constraint-Based Routing and is the key to MPLS traffic engineering. Constraint-Based Routing manages traffic paths within an MPLS network, allowing traffic to be steered to desired paths.

MPLS traffic engineering also enables resiliency and reliability to be built into carrier networks, increasing the availability and value of the network to their customers. Using MPLS traffic engineering, LSP connections can be optimized and preempted. When outages occur, traffic can be actively rerouted around failed links. An example of this is RSVP-TE Fast Reroute, which provides for sub-50ms switchovers between primary and back up LSPs or LSP bundles.

Traffic engineering is deployed in MPLS networks via traffic engineering extensions to IGPs, such as OSPF and IS-IS. OSPF-TE and IS-IS-TE carry additional information ?such as link bandwidth, link utilization, delay, priority, preemption, etc. ?to allow the network to utilize paths that meet service requirements, resource availability, load balancing, and failure recovery objectives. RSVP-TE is widely used for MPLS signaling in networks that require traffic engineering.

MPLS traffic engineering is typically deployed in the core of the MPLS network, while QoS is used at the edge. QoS at the edge ensures that high priority packets get preferential treatment, while traffic engineering avoids network congestion and appropriately utilizes available bandwidth resources. Together, QoS and traffic engineering enable organizations to move away from multiple, specialized networks for voice, video, and data to a single converged IP/MPLS network, dramatically reducing overhead and cost.

MPLS Challenges

MPLS has made significant progress over the last few years and is well into mainstream deployment in networks around the world. But key challenges to attaining more widespread acceptance remain. MPLS encompasses a wide range of functionality and applications, therefore its implementation has an associated high level of complexity. Vendors who develop MPLS technology, as well as organizations looking at deploying MPLS in a network today, must also factor in MPLS抯 continually evolving state and its impact on network performance and scalability.

MPLS is not a standalone technology ?it is overlaid on Layer 2 technologies such as Ethernet or ATM, and must operate in conjunction with other control plane protocols, such as IP routing. The complexity of MPLS deployments is increased because of this interaction. In some cases, four or more protocols may be involved in a given network scenario, necessitating careful coordination and validation of the end-to-end system. Integration of legacy services and deployment of new services, such as VPNs, requires tunneling, which in turn increases the setup requirements for a given circuit.

Though under development for a number of years, MPLS continues to evolve rapidly. The primary goals of the technology have shifted over time as technology has progressed. Today, a number of extensions to the MPLS protocols, as well as new functionality, are under development. New developments often obsolete older ones. This dynamic nature presents a moving target to those developing and deploying MPLS. Vendors must decide whether to implement a new feature with an eye on the industry抯 current direction. Service providers gauge the viability of new developments by asking whether they solve a given problem better. On certain issues, the industry has split into multiple camps, further complicating the situation as organizations trade off the long term risk of obsolescence with the shorter term benefits of implementing a certain technology. Interoperability of MPLS equipment in heterogeneous networks remains an issue and will continue to be so for several years to come.

Though advances in silicon technology have vastly improved the raw performance of today抯 routers, the complexity of MPLS in real network applications presents performance and scalability issues. The challenge is typically not in the MPLS core network, where data is simply being label-switched, but at the network edge where MPLS must integrate with non-MPLS networks, and where services are initiated. As networks converge, traffic loads increase and networks must be able to deal with the overhead of handling real-time and prioritized traffic. The meshed connections required for VPN deployments can quickly challenge equipment scalability limitations, as well as provisioning and management requirements. Large service provider networks have the ultimate scalability challenge and must consider the limitations of their equipment as they look to maximize return on investment.

The challenges presented by MPLS ultimately necessitate thorough testing and validation of MPLS systems during development and prior to deployment. MPLS is by no means a plug-and-play proposition. Performing appropriate due diligence is the only way to ensure success when dealing with a broad-scope and rapidly evolving technology such as MPLS.

标签：

版权申明：本站文章部分自网络，如有侵权，请联系：west999com@outlook.com
特别注意：本站所有转载文章言论不代表本站观点，本站所提供的摄影照片，插画，设计作品，如需使用，请与原作者联系，版权归原作者所有