![Huawei Quidway NetEngine40E Configuration Manual - Reliability Download Page 21](http://html.mh-extra.com/html/huawei/quidway-netengine40e/quidway-netengine40e_configuration-manual-reliability_169453021.webp)
l
Hierarchical networking: A network is divided into three layers, core layer, convergence
layer, and edge layer. According to the current service status and future service prediction,
redundancy backup is required on a device connected to access nodes on the edge layer.
The active and standby nodes connect to convergence nodes. Devices of convergence layer
are dual-homed to single node multi-device of the upper layer or to multi-node device of
convergence layer and core layer alternatively. Devices of core layer are enabled with full
interconnection or half interconnection. In this manner, two devices are reachable to each
other with one route at a fast traffic rate, avoiding multi-interconnection.
l
On the same layer, multi-interconnection is recommended; multi-device is applicable to a
single node.
l
The lower-layer devices are dual-homed or multi-homed to single nodes or multiple nodes
of devices on the upper layer.
l
Adjustment should be taken according to traffic.
1.2 Reliability Technologies for IP Networks
This section describes two typical reliability technologies, failure detection and switchover.
1.2.1 Failure Detection for IP Networks
Fault detection technologies are divided into special detection technologies and common
detection technologies.
l
Special fault detection technologies include:
–
Automatic Protection Switching (at the transport layer)
–
Eth-OAM (at the link layer)
–
MPLS (Multiprotocol Label Switching) OAM (for MPLS)
l
Common fault detection technologies include the Bidirectional Forwarding Detection
(BFD) that detects faults at all layers, such as the data link layer and the network layer.
The fault detection mechanism is available to each layer of the TCP/IP reference module,
including:
l
Transport/Physical layer: Automatic Protection Switching (APS)
l
Data link layer: MPLS OAM, Eth-OAM, STP, RSTP, MSTP, and RRPP
l
Network layer: Hello mechanism for protocols, VRRP (Virtual Router Redundancy
Protocol), and GR (Graceful Restart)
l
Application layer: Heartbeat mechanism and retransmission mechanism for protocols
Fault detection has the following modes:
l
Asynchronous mode: Detection packets are sent periodically.
l
Query mode: A series of packets for confirmation are sent.
l
Echo mode: The received packet is sent back to the sender without any change.
1.2.2 Protection Switching for IP Networks
The standard time of protection switching on a data communication network is 50 ms. Link
redundancy is a basis of protection switching.
HUAWEI NetEngine80E/40E Router
Configuration Guide - Reliability
1 Reliability Overview
Issue 02 (2014-09-30)
Huawei Proprietary and Confidential
Copyright © Huawei Technologies Co., Ltd.
4