Browse Prior Art Database

MPLS Tunnel Switching based on performance measurements

IP.com Disclosure Number: IPCOM000131945D
Published in the IP.com Journal: Volume 5 Issue 12A (2005-12-25)
Included in the Prior Art Database: 2005-Dec-25
Document File: 2 page(s) / 318K

Publishing Venue

Siemens

Related People

Juergen Carstens: CONTACT

Abstract

In computer networking and telecommunications, the Multiprotocol Label Switching (MPLS) is a data-carrying mechanism, operating at a layer below protocols such as IP (Internet Protocol). It can be used to carry many different kinds of traffic, including both voice telephone traffic and IP packets. In case of Local Protection of a network using MPLS-TE (TE, Traffic Engineering) each node or link which has to be protected implies an individual Backup tunnel. If a loss of connectivity in the Main tunnel is detected by the control plane, the traffic is redirected to the Backup tunnel as shown in Figure 1. In case the MPLS LSP (Label Switching Path) is failing to deliver data traffic, it is not always possible to detect the failure using the MPLS control plane. For instance, the control plane of the MPLS LSP can be functional while the data plane is forwarding incorrect or fragmentary data. Therefore, the link conditions can be degraded in the manner that a loss of connectivity is not detected by the control plane, but in the data plane the quality cannot be maintained. In that case the protection mechanisms are not activated and the traffic is getting lost and the quality of service (QoS) is compromised. For failures detected by the MPLS control plane there are two protection mechanisms with proposed standards in the IETF (Internet Engineering Task Force): Global Protection (IETF Rfc 3469) and Local Protection (IETF Rfc 4090). These mechanisms only consider the situation of complete link failure and all packets that are already in the tunnel are lost. A disadvantage of this mechanism is the need of complete loss of connectivity in order to switch to the Backup tunnel. Concerning failures in the data plane, there are some proposals posted as drafts in the IETF which detect this kind of failure but the detection is not used to initialize the switching to the Backup tunnel.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

S

MPLS Tunnel Switching based on performance measurements

Idea: Eduardo Mendes, PT-Lisboa; Pedro Charrua, PT-Lisboa

In computer networking and telecommunications, the Multiprotocol Label Switching (MPLS) is a data- carrying mechanism, operating at a layer below protocols such as IP (Internet Protocol). It can be used to carry many different kinds of traffic, including both voice telephone traffic and IP packets. In case of Local Protection of a network using MPLS-TE (TE, Traffic Engineering) each node or link which has to be protected implies an individual Backup tunnel. If a loss of connectivity in the Main tunnel is detected by the control plane, the traffic is redirected to the Backup tunnel as shown in Figure 1. In case the MPLS LSP (Label Switching Path) is failing to deliver data traffic, it is not always possible to detect the failure using the MPLS control plane. For instance, the control plane of the MPLS LSP can be functional while the data plane is forwarding incorrect or fragmentary data. Therefore, the link conditions can be degraded in the manner that a loss of connectivity is not detected by the control plane, but in the data plane the quality cannot be maintained. In that case the protection mechanisms are not activated and the traffic is getting lost and the quality of service (QoS) is compromised.

For failures detected by the MPLS control plane there are two protection mechanisms with proposed standards in the IETF (Internet Engineering Task Force): Global Protection (IETF Rfc 3469) and Local Protection (IETF Rfc 4090). These mechanisms only consider the situation of complete link failure and all packets that are already in the tunnel are lost. A disadvantage of this mechanism is the need of complete loss of connectivity in order to switch to the Backup tunnel. Concerning failures in the data plane, there are some proposals posted as drafts in the IETF which detect this kind of failure but the detection is not used to initialize the switching to the Backup tunnel.

In the following, it is proposed to switch between the Main and Backup tunnel not only in case of loss connectivity, but also in case of serious degradation of link transmission conditions. Based on delay and jitter measurements made in the data plane of the Main and Backup tunnel, the decision to switch to the Backup Tunnel is made before any loss of connectivity occurs, shown in Figure 2.

In order to implement the idea, a new switching mechanism is implemented as an extension to the existent protection mechanisms.

The following components are implemented in the LSRs (LSR, Label Switching Router):

1) Decision Point

A...