Reliability Principles of RAN Networks (Applicable for the GBSS15.0) - 20130115-A-1.0

5/22/2018 Reliability Principles of RAN Networks (Applicable for the GBSS15.0) - 20130115-A-1.0

1/16

www.huawei.com

Copyright 2013 Huawei Technologies Co., Ltd. All rights reserved.

Reliability Principles

of RAN Networks


2/16Copyright 2013 Huawei Technologies Co., Ltd. All rights reserved. Page 2

Overview

This document describes network design for Huawei

BSC6900/6910 with an emphasis on reliability principles on

the RAN.


3/16Copyright 2013 Huawei Technologies Co., Ltd. All rights reserved. Page 3

Contents

1. Active/Standby Boards with Manual Active/Standby LAGs

2. Active/Standby Boards with Dual Active Ports

3. Standalone Board with Load-Sharing Ports


4/16Copyright 2013 Huawei Technologies Co., Ltd. All rights reserved.

Basic Principles of IP Networking

DEV IP: Specifies the logical IP address for BSC board communication. This logical IP address is shared by the

active and standby boards.

Port IP: IP1 is the IP address of the active port. The standby port needs to configure the IP address only when an

ARP detection is needed.

Bearer network: L3 routers are Router 1 and Router 2. These two routers use VRRP to enhance router.

VRRP IP9: Specifies the virtual IP addresses of Router 1 and Router 2.

Routing: Specifies the path from the active BSC to the MGW. The path is IP1----VRRP IP----IP8, in which IP1 is the

source address, VRRP IP9 is the MGW address, and IP8 is the destination address. Because Router 1 and Router 2

use VRRP, VRRP IP9 replaces IP3 and IP4 as the actual MGW address.

MGW

Router 3

Router 4

MSC

Server

Router 1

Router 2

Active Standby

DEV IP

IP 1

V

R

R

P

IP 3

IP 4

VRRP IP 9

IP 6 IP 7

IP 8

BSC

IP 5

Page 4



Principles of Manual Active/Standby LAGs

Two FE/GE links on a board on the BSC6900 side

are configured into an aggregation group that

connects to two routers. The aggregating mode of the aggregation group

must be manual mode. The working mode of the

aggregation group must be active/standby mode.

Two FE/GE links on the active and standby

boards are configured into an aggregation

group that connects to VRRP devices.

Page 5



Active/Standby Boards with Manual

Active/Standby LAGs

Link aggregation applies to EXOUa (BSC6910)/FG2d/FG2c/GOUd/GOUc interface boards.

Aggregation groups must be created in slots in the logical board. The TRUNK group supports bearing logical ports.

An aggregation group may consist of multiple sub-links. A member port must be allocated in the active/standby boards

which contain slots of aggregation ports. Member ports of the TRUNK group have the same data rate and work in full

duplex mode. The ports are not configured with IP addresses, logical ports, or ETHMEP. For EXOUa boards, the maximumnumber of supported 10GE ports in an aggregation group is 2. For FG2d/FG2c/GOUd/GOUc boards, the maximum

number of supported member ports in an aggregation group is 8.

The number of IP addresses configured in each aggregation group is up to 6.

MGW

Router 3

Router 4

MSC

Server

Router 1

Router 2

Active Standby

DEV IP

IP 1

V

R

R

P

IP 3

IP 4

VRRP IP 9

IP 6 IP 7

IP 8

BSC

IP 5

Page 6



Principles of the IP Pool

In a transmission pool, the peer NE can simultaneously connect

to multiple interface boards on the BSC side. If an interface board

on the BSC side is faulty, services will not be interrupted. This

improves the network reliability.

This solution requires a three-layer networking between the BSC

and the MGW to ensure full mesh between the MGW and all

interface boards on the BSC side.

After multiple transmission interface boards on the BSC side form

a transmission resource pool, the MGW can communicate with

any interface board in the pool. When the load of an interface

board in the pool is high, the BSS automatically distributes

subsequent calls to interface boards with light traffic. In addition,

the BSC implements the free configurations on IP routes and IP

paths. That is, when the IP address is added or modified on the

MGW, the BSC does not need to change the configuration of the

corresponding IP path or IP route.

Page 7



IP Interface Boards - Independent Switchover of Different

Manual Active/Standby LAGs (Recommended)

After a

switchover

The solution of active/standby boards with manual active/standby LAGs is based on port aggregation. In this

mode, the active port sends and receives packets, while the standby port is normally unused or is only used

in link detection. Active/standby ports are separate from active/standby boards. The active port can be oneither the active or standby board.

As shown above, after a switchover, active and standby boards remain unchanged. Port 0 of the active

and standby boards are switched over. Port 0 uses IP1 in the standby board.

The networking based on active/standby boards, active/standby ports, Layer 3 networking, router VRRP,

and device IP for communication are recommended.

Router 1

Router 2

Active Standby

DEV IP

Port 0

V

R

R

P

IP 3

IP 4

VRRP IP 9

VRRP IP 10

BSC

ort 0 : IP 1

Port 1 :IP 2Port 1

VR

R

P

Router 1

Router 2

Active Standby

DEV IP

Port 0 : IP 1

VR

R

P

IP 3

IP 4

VRRP IP 9

VRRP IP 10

BSC

Port 0

Port 1 : IP2Port 1

VR

R

P

Page 8


9/16


IP Interface Boards - Principles of

Active/Standby Board Switchover

After a

switchover

After the active and standby boards are switched over, their active/standby status ischanged. Logical IP address moves with the active board. Port IP addresses move

with the active port. Difference between manual active/standby LAGs and

active/standby ports is not about the networking but about the command

configuration of the internal MML.

Router 1

Router 2

Active Standby

DEV IP

Port 0

V

R

R

P

IP 3

IP 4

VRRP IP 9

BSC

Port 0 : IP 1

Port1 : IP 2Port 1

Router 1

Router 2

ActiveStandby

DEV IP

Port 0 : IP 1

V

R

R

P

IP 3

IP 4

VRRP IP 9

BSC

Port 0

Port 1Port1 : IP 2

Page 9


10/16


Contents




Page 10


11/16


IP Networking - Active/Standby Boards with

Dual Active Ports (Optional)

Router 1

Router 2

Active Standby

DEV IP

IP 5

IP 6BSC

IP 1 IP 2

IP 3IP 4

After a

switchover

Features: The DEV IP is configured on the logical active board. The two ports use IP addresses of different network segments.

Based on the active/standby MGW features of the source IP, the active and standby paths are bound to the egress ports ofthe active and standby boards to achieve active/standby routes of the active and standby boards. In addition, the DEV IP is

used to configure the egress ports on the active and standby boards and configure active and standby routes on the router, so

that the ports of the active and standby boards can provide backup for each other. After the active/standby boards are

switched over, routes will be switched over but ports will keep unchanged. Compared with the solution of active/standby

boards with active/standby ports, the solution of active/standby boards with dual active ports provides a twice capacity due to

the load sharing on both active ports but requires a more complex configuration.

Router 1

Router 2

Standby Active

DEV IP

IP 5

IP 6BSC

IP 1 IP 2

IP 3IP 4

Page 11


12/16


Contents




Page 12


13/16


IP Interface Boards - Standalone Board with Load-

Sharing Ports (Optional)

Two boards work independently, achieving inter- and intra-board load sharing. Priority level ofroutes from the two boards to the destination address must be the same.

Characteristics: External load balancing between boards is enabled without board

active/standby configuration. If one board is faulty, ongoing calls are disconnected, but new

calls can still be accepted.

Router 1

Independent

IP Backbone

IP 2 IP 3

BSC

IP 1

Independent

IP 4

Page 13


14/16


Detection Mechanism: BFD and ARP

Recommended detection methods:

1.The active port uses single-hop dual BFDs for fault detection. The destination IP addresses are the port IP addresses of the two VRRP routers in the

next-hop (IP3 and IP4 shown above).

2.The standby port uses the ARP for fault detection. The destination IP address is the VRRP IP1 of the router. The standby port must be configured with

an IP address to initiate the ARP detection. (Configure the IP address for the standby port when configuring the ARP detection rather than when running

ADD ETHIP.)

3.If faults are found on the active port by both BFDs but no fault is found on the standby port by the ARP, an active/standby ports switchover is triggered.

4.Networking solutions using the BFD and ARP for fault detection are as follows:

- Active/standby boards with manual active/standby LAGs;

- Active/standby boards with dual ports.

If the A interface uses the IP pool, enable the ICMP ping detection in the IP pool.

Router 1

Router 2

Active Standby

DEV IP

IP 2

V

R

R

P

IP 3

IP 4

VRRP IP 9

BSC

IP 1

BFD to router1

BFD to router2

ARP to

Router2

Page 14


15/16

Copyright 2013 Huawei Technologies Co., Ltd. All rights reserved. Page 15

OM Networking Design

Physical IP2:

10.161.26.22/

29

Physical IP1:

10.161.26.21/

29

Logical IP:

10.161.26.23/29

Router 1

Router 2

SW1

SW2

V

R

R

P

TRUNK

For dual OMUs, three IP addresses are required: physical IP addresses for eachOMU and a shared logical IP address. The three IP addresses must be on the same

network segment.

Port switchover within the same board is preferred.


16/16

www.huawei.com

Thank you

Documents

Reliability Principles of RAN Networks (Applicable for the GBSS15.0) - 20130115-A-1.0