How to improve the reliability of network layer of metropolitan Area Network router

The rapid development of broadband business has brought about profound changes in traditional telecom industry and it business. The integration of multiple services and multiple networks has become an irreversible trend. As the main network entity in metropolitan Area Network, the broadband metropolitan area Network will become the host platform of 3G, NGN and other new value-added services. 3G, NGN, such as real-time voice and video applications, requires metropolitan area network to provide quality of service assurance and similar to traditional telecommunications technology 99.999% of telecommunications-grade network reliability requirements. At the same time, the intense competition also impels the operator to provide the customer with the similar SLA to have the service Quality Assurance service, the network reliability is one of the first and most important indicators. By improving the reliability of the network, operators can provide differentiated services, occupy a favorable position in the competition among operators, and further establish and consolidate the brand image of enterprises.

The reliability of metropolitan area Network router is embodied in the following two aspects, one is the reliability of the equipment layer, the other is the reliability of the network layer.

The network reliability is an important content of the reliability guarantee of metropolitan area Network router, because the traditional router protocol converges slowly (IGP in the second level, BGP in the minute level), can not meet the demand of carrying real time service. Network reliability is also an active field of metropolitan area Network router technology.

At present, the new network layer reliability technology includes fast convergence of IP routing, end-to-end LSP backup, MPLS fast rerouting, smooth restart, rprips and so on.

Fast convergence of IP routing

IP dynamic routing is the most basic network layer reliability guarantee mechanism, and is the inherent function of IP routing network. IP Dynamic routing protocol is responsible for network layer IP Forwarding path calculation, the routing protocol dynamically recalculates the data forwarding path when the link or node failure causes the original data forwarding path to be interrupted, although the response time of various routing protocols varies with different mechanisms, but the average level is at the second level. The recovery time of traditional IP service is acceptable, but it requires the millisecond recovery response time for telecom-grade IP network which can carry the real-time service and so on, the traditional IP dynamic routing technology and this request have a big gap.

On the basis of traditional routing protocol, improvement can shorten the fault response time of IP routing protocol, which is mainly to speed up the convergence of routing protocol. The speed of convergence of routing protocols can be divided into several aspects, such as link fault detection, route weight calculation and routing information updating. By speeding up the transmission frequency of the Hello message between the links, speeding up the SPF calculation speed and setting high priority for the route update message, the routing protocol can quickly discover and handle the fault, and update the route accurately and quickly, accelerate the convergence of the routing protocol, and realize the convergence of less than 1s by optimizing the IGP routing

Another way to speed up the convergence of routing protocols is to use IGP and EGP to plan the network rationally, IGP the routing of the equipment in the domain, the EGP (BGP4) carries the external route, and the two routes are effectively isolated and not redistributed to each other. The reasonable division of IGP and BGP has formed a hierarchical routing structure, and the convergence of routing protocols between domains and domains is independent and mutually exclusive, and can achieve the fastest convergent speed.

LSP Protection Switching

Protection switching is the term used by ITU-T, and protection switching technology is critical to improve the availability and stability of MPLS networks. Protection switching generally computes and allocates resources to a protected LSP route, so you can quickly regain network resources after an LSP connection fails or is interrupted.

The current technology development can only support the point-to-point LSP protection switch, protection can take two ways: 1+1 protection and 1:1 protection.

1+1 Protection uses a dedicated backup LSP as the primary LSP protection, where the primary LSP and the backup LSP Bridge are connected, the traffic on the primary LSP is replicated to the backup LSP and delivered to the EGRESSLSR,EGRESSLSR based on the value of the fault indication parameter, Select to receive traffic on the main standby LSP.

1:1 Protection also uses a dedicated backup LSP as the primary LSP protection, but the primary LSP does not transmit the same traffic at the same time, the backup LSP can transmit other traffic while the main LSP works normally, and the traffic protection switching decision is INGRESSLSR.

MPLS Fast Reroute (FRR)

In order to meet real-time applications such as video conferencing TV, these traffic must be provided with a traditional sdhaps-millisecond LSP protection capability.

The LSP protection switching technology requires the intervention of signaling protocol, and the fault signal transmission of fault point to recovery point introduces unnecessary network recovery delay. The MPLS fast rerouting technology can realize that the fault link traffic is redirected according to the preset protection path without signaling intervention, and the recovery point is the fault point. Most fast rerouting schemes rely on a pre-established backup channel, and when a network recovery point detects a network failure, it is simply updating the LSP Interchange table, switching traffic from the faulty port LSP to the LSP built in advance on the normal port.

The advantage of fast rerouting, in addition to improving the speed of protection recovery, is to configure the protection capability through selective network vulnerabilities, avoiding redundant protection in reliable networks and unnecessary consumption of core network resources. MPLS Fast rerouting technology provides protection switching within 50ms, and can be used as an alternative to sdhaps protection mechanism. (☆ Introduction to Programming Network ☆)

MPLS Fast Reroute uses the following configuration process:

First of all, at the entrance of LSP is LSR1, using a user command to activate the MPLS protection switching function; LSR1 sends a signaling to all LSR on the LSP path, each LSR calculates a bypass next hop LSR backup LSP,LSP Fast reroute configuration is complete. When one of the LSR on the LSP path detects a downstream failure, the LSR switches the traffic locally to the backup LSP.

There are many kinds of fast rerouting schemes in the IETF, the two main protection modes are link protection and node protection, and their solutions and complexity are different, and the technology has not yet formed a formal RFC.

Smooth reboot (Gracefulrestart)

The possible factors that cause the control plane reboot include: software upgrade, software bug or hardware failure, nondisruptive Restart can be achieved when the control plane restart, the data plane without interruption forwarding. However, if the control plane failure, the peer router will recalculate the route, bypass the fault router, the data plane of the uninterrupted forwarding is meaningless, and the fault route will spread to the entire network scope. If this happens on the MPLSVPNPE router, the result is catastrophic.

Control plane Smooth Restart technology can effectively solve this problem, the router using this technology in the control plane failure, you can notify neighboring routers continue to use the original path for data forwarding, while restarting the router and neighboring routers to establish a routing status, to ensure the restart process of business availability, Minimize the impact of a single device reboot on the entire network.

The router does not save the associated protocol state during a smooth reboot, so the restart software failure caused by the reboot will not continue after the reboot.

A smooth reboot is a new feature that many old devices cannot support, so it can be used on devices that support this feature on a local subnet.

In the network boundary, the operator boundary router faces many customers, and generally has no redundancy measures, which is most suitable for using the smooth restart technology. Network core is generally used to protect the redundant path, and with business restart is easy to cause routing ring, so it is not recommended to use a smooth restart in the network core technology.