FusionStorage

8.0.1

Block Storage HyperReplication

Feature Guide

Issue 05
Date 2021-02-05

HUAWEI TECHNOLOGIES CO., LTD.

Copyright © Huawei Technologies Co., Ltd. 2021. All rights reserved.
No part of this document may be reproduced or transmitted in any form or by any means without prior
written consent of Huawei Technologies Co., Ltd.

Trademarks and Permissions

and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.
All other trademarks and trade names mentioned in this document are the property of their respective
holders.

Notice
The purchased products, services and features are stipulated by the contract made between Huawei and
the customer. All or part of the products, services and features described in this document may not be
within the purchase scope or the usage scope. Unless otherwise specified in the contract, all statements,
information, and recommendations in this document are provided "AS IS" without warranties, guarantees
or representations of any kind, either express or implied.

The information in this document is subject to change without notice. Every effort has been made in the
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute a warranty of any kind, express or implied.

Huawei Technologies Co., Ltd.

Address: Huawei Industrial Base
Bantian, Longgang
Shenzhen 518129
People's Republic of China

Website: https://e.huawei.com

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. i

FusionStorage
Block Storage HyperReplication Feature Guide About This Document

About This Document

Purpose
HyperReplication is the remote replication feature in FusionStorage storage
systems developed by Huawei. This document describes its working principles,
application scenarios, planning and preparation, software installation, as well as
configuration and management.

Intended Audience
This document is intended for:

● Technical support engineers

● Maintenance engineers

Symbol Conventions
The symbols that may be found in this document are defined as follows.

Symbol Description

Indicates a hazard with a high level of risk which, if not

avoided, will result in death or serious injury.

Indicates a hazard with a medium level of risk which, if not

avoided, could result in death or serious injury.

Indicates a hazard with a low level of risk which, if not

avoided, could result in minor or moderate injury.

Indicates a potentially hazardous situation which, if not

avoided, could result in equipment damage, data loss,
performance deterioration, or unanticipated results.
NOTICE is used to address practices not related to personal
injury.

Supplements the important information in the main text.

NOTE is used to address information not related to personal
injury, equipment damage, and environment deterioration.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. ii

FusionStorage
Block Storage HyperReplication Feature Guide About This Document

Change History
Issue Date Description

05 2021-02-05 This issue is the fifth official release.

Content has been optimized.

04 2020-11-30 This issue is the fourth official release.

Content has been optimized.

03 2020-08-31 This issue is the third official release.

Content has been optimized.

02 2020-03-01 This issue is the second official release.

Content has been optimized.

01 2019-10-31 This issue is the first official release.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. iii

FusionStorage
Block Storage HyperReplication Feature Guide Contents

Contents

About This Document................................................................................................................ ii

1 Feature Description................................................................................................................. 1
1.1 Overview.................................................................................................................................................................................... 1
1.2 Availability................................................................................................................................................................................. 2
1.3 Working Principles.................................................................................................................................................................. 2
1.3.1 Basic Concepts...................................................................................................................................................................... 2
1.3.2 Data Replication................................................................................................................................................................... 3
1.3.3 Functions of Consistency Groups....................................................................................................................................5

2 Planning and Preparation......................................................................................................8

2.1 Typical Networking................................................................................................................................................................. 8
2.1.1 Network Concepts............................................................................................................................................................... 8
2.1.2 Deployment Solutions......................................................................................................................................................12
2.2 System Requirements.......................................................................................................................................................... 33
2.3 Required Software Packages and Documentation.................................................................................................... 34

3 Software Installation............................................................................................................36
4 Configuring HyperReplication............................................................................................37
4.1 Configuration Process.......................................................................................................................................................... 37
4.2 Checking the License........................................................................................................................................................... 38
4.3 Configuring Basic Services................................................................................................................................................. 39
4.4 Creating a Replication Cluster.......................................................................................................................................... 39
4.5 Adding a Remote Device.................................................................................................................................................... 47
4.6 Creating a Remote Replication Pair............................................................................................................................... 48
4.7 Creating a Remote Replication Consistency Group.................................................................................................. 52

5 Managing HyperReplication............................................................................................... 57
5.1 Managing a Remote Replication Pair............................................................................................................................ 57
5.1.1 Viewing Information About a Remote Replication Pair....................................................................................... 57
5.1.2 Modifying Properties of a Remote Replication Pair.............................................................................................. 61
5.1.3 Synchronizing a Remote Replication Pair................................................................................................................. 63
5.1.4 Splitting a Remote Replication Pair............................................................................................................................ 64
5.1.5 Performing a Primary/Secondary Switchover.......................................................................................................... 65
5.1.6 Enabling Protection for Secondary Resources......................................................................................................... 66

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. iv

FusionStorage
Block Storage HyperReplication Feature Guide Contents

5.1.7 Disabling Protection for Secondary Resources........................................................................................................ 67

5.1.8 Deleting a Remote Replication Pair............................................................................................................................ 68
5.2 Managing a Remote Replication Consistency Group............................................................................................... 69
5.2.1 Viewing Remote Replication Consistency Group Information........................................................................... 69
5.2.2 Modifying Properties of a Remote Replication Consistency Group................................................................. 72
5.2.3 Synchronizing a Remote Replication Consistency Group.................................................................................... 74
5.2.4 Splitting a Remote Replication Consistency Group............................................................................................... 75
5.2.5 Performing a Primary/Secondary Switchover for a Remote Replication Consistency Group..................76
5.2.6 Enabling Protection for Secondary Resources in a Remote Replication Consistency Group...................77
5.2.7 Disabling Protection for Secondary Resources in a Remote Replication Consistency Group................. 78
5.2.8 Adding a Remote Replication Pair............................................................................................................................... 79
5.2.9 Removing a Remote Replication Pair......................................................................................................................... 80
5.2.10 Deleting a Remote Replication Consistency Group.............................................................................................80
5.3 Managing a Replication Cluster.......................................................................................................................................81
5.3.1 Expanding a Replication Cluster...................................................................................................................................82
5.3.2 Reducing a Replication Cluster..................................................................................................................................... 83
5.3.3 Removing a Remote Device........................................................................................................................................... 83
5.3.4 Configuring a Cluster Service Credential...................................................................................................................83
5.3.5 Changing a Replication Cluster Name....................................................................................................................... 84
5.3.6 Deleting a Replication Cluster...................................................................................................................................... 85

A Appendix................................................................................................................................. 86
A.1 Modifying the RoCE Configuration File........................................................................................................................ 86
A.2 Configuring the Replication Network by Modifying the Configuration File.................................................... 87

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. v

FusionStorage
Block Storage HyperReplication Feature Guide 1 Feature Description

1 Feature Description

HyperReplication is the remote replication feature developed by Huawei. It

provides flexible and powerful data replication functions to achieve data backup
and recovery, continuous support for service data, and disaster recovery (DR).
1.1 Overview
1.2 Availability
1.3 Working Principles

1.1 Overview
This section describes the background, definition, and benefits of the
HyperReplication feature.

Background
As the digitalization drive advances in various industries, data has become critical
to the operation of enterprises, and customers impose increasingly demanding
requirements on the stability of storage systems. Although some storage devices
offer extremely high stability, they fail to prevent irrecoverable damage to
production systems in the event of a natural disaster.

To ensure continuity, recoverability, and high reliability of service data, remote DR

solutions emerge. The remote replication technology is one of the key
technologies used in remote DR solutions.

Definition
HyperReplication implements asynchronous remote replication to periodically
synchronize data between the primary and secondary storage systems to support
system DR. This minimizes service performance deterioration caused by the
latency of long-distance data transmission.

Benefits
The HyperReplication feature provides the following benefits:

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 1

FusionStorage
Block Storage HyperReplication Feature Guide 1 Feature Description

● Secondary backup and recovery

Service data can be recovered using backup data at the secondary site after
the service data becomes unusable. This prevents damage caused by data loss
in the event that data at the primary site becomes unusable.
● Continuous support for services
Service data can be quickly switched from the primary site to the secondary
site, protecting service continuity. This prevents losses caused by a service
interruption upon a failure at the primary site.
● DR
Data at the primary site can be recovered using backup data at the secondary
site in the event of a disaster. This prevents problems that would otherwise be
caused by service data loss or a long recovery duration in the event of a
disaster.

1.2 Availability
License Requirement
The HyperReplication feature requires a standard license or an advanced license.
Table 1-1 shows their application scenarios.

Table 1-1 Application scenarios of HyperReplication licenses

Standard Edition Advanced Edition

RPO ≥ 1 minute RPO ≥ 10 seconds

Version Support
Both the primary and secondary storage systems must be FusionStorage 8.0.0 or
later.

1.3 Working Principles

This section describes the basic concepts and data replication principles of the
HyperReplication feature, as well as introduces the functions of consistency
groups.

1.3.1 Basic Concepts

Primary Volume and Secondary Volume

Volumes in a primary storage system are called primary volumes, and volumes in
a secondary storage system are called secondary volumes.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 2

FusionStorage
Block Storage HyperReplication Feature Guide 1 Feature Description

Pair
A pair is the data replication relationship between a primary volume and a
secondary volume. One primary volume in the primary storage system and one
secondary volume in the secondary storage system form a pair.

Consistency Group
A consistency group is a collection of pairs that have a service relationship with
each other. For example, the primary storage system has three primary volumes
that respectively store service data, logs, and change records of a database. If data
on any of the three volumes becomes invalid, the data on all three volumes
becomes unusable. For the three pairs in which these three volumes exist, you can
create a consistency group. In actual configuration, you need to create a
consistency group and then manually add the three pairs to the consistency group.

Synchronization
Synchronization is the process of copying data from a primary volume to a
secondary volume.

Splitting
Splitting is the process of suspending the remote replication relationship between
primary and secondary volumes.

Primary/Secondary Switchover
A primary/secondary switchover is the process of exchanging the roles of the
primary and secondary volumes in a pair.

Secondary Resource Protection

Secondary volumes are read-only.

Replication Cluster
A replication cluster consists of a replication control cluster and a replication
service cluster:
● A replication control cluster manages cluster nodes and metadata and
contains three, five, seven, or nine nodes. The CCDB process of the replication
control cluster requires local storage space to record metadata information.
● A replication service cluster manages remote replication pairs and consistency
groups and contains 3 to 64 nodes.

1.3.2 Data Replication

The HyperReplication feature periodically synchronizes data between primary and
secondary sites to implement system DR. This minimizes service performance
deterioration caused by remote data transmission latency.
Each remote replication pair consists of a primary volume and a secondary
volume. Figure 1-1 shows the data replication principle.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 3

FusionStorage
Block Storage HyperReplication Feature Guide 1 Feature Description

Figure 1-1 Data replication principle

1. In normal cases, the primary volume carries data read and write services.
Before data synchronization, data at the primary and secondary sites is
inconsistent.
2. After data synchronization, data at the primary and secondary sites is
consistent. The following describes data replication when the primary volume
becomes faulty.

NOTICE

● Secondary resource protection must be enabled before data

synchronization.
● Secondary resource protection must be disabled before a primary/
secondary switchover.

– If the primary volume becomes faulty and data at the primary and
secondary sites is consistent:
i. Disable secondary resource protection. The secondary volume takes
over data read and write services.
ii. After the primary volume recovers, perform a primary/secondary
switchover to designate the secondary volume as the primary one.
Then enable secondary resource protection and synchronize data.
– If the primary volume becomes faulty and data at the primary and
secondary sites is inconsistent:
i. Disable secondary resource protection. The secondary volume takes
over data read and write services.
ii. After the primary volume recovers, perform a primary/secondary
switchover to designate the secondary volume as the primary one.
Then enable secondary resource protection and synchronize data.
Data to be synchronized from the original primary volume to the
original secondary volume will be lost.
– If the primary volume becomes faulty:

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 4

FusionStorage
Block Storage HyperReplication Feature Guide 1 Feature Description

i. The secondary volume directly takes over data read and write
services.
ii. After the primary volume recovers, it takes back the data read and
write services. Data at the primary and secondary sites is
inconsistent.
iii. Enable secondary resource protection and synchronize data. Data
written to the secondary volume when the secondary volume carries
services will be lost.

1.3.3 Functions of Consistency Groups

In medium- and large-sized database applications, data, logs, and change records
are stored in associated volumes of storage systems. If data of one volume is lost,
data in other volumes becomes invalid. To ensure data availability, you can add
remote replication pairs of these volumes to the same consistency group.
This section compares storage systems running consistency groups with storage
systems not running consistency groups to illustrate how consistency groups help
prevent data loss.

Remote Replication Without Using a Consistency Group

Associated volumes that are not added to a consistency group are at a risk of
becoming unavailable, as shown in Figure 1-2.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 5

FusionStorage
Block Storage HyperReplication Feature Guide 1 Feature Description

Figure 1-2 Remote replication without using a consistency group

Remote Replication Using a Consistency Group

If associated volumes are added to a consistency group, data unavailability is
prevented, as shown in Figure 1-3.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 6

FusionStorage
Block Storage HyperReplication Feature Guide 1 Feature Description

Figure 1-3 Remote replication using a consistency group

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 7

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

2 Planning and Preparation

Before the deployment, plan each network and ensure that the hardware meets
the requirements. Prepare required software packages and tools.
Before configuring HyperReplication, you can plan related data in advance
according to FusionStorage Block Storage LLD Configuration Template.
2.1 Typical Networking
This section describes network concepts and deployment solutions of
HyperReplication.
2.2 System Requirements
2.3 Required Software Packages and Documentation

2.1 Typical Networking

This section describes network concepts and deployment solutions of
HyperReplication.

2.1.1 Network Concepts

FusionStorage consists of a management network, service network, storage
network, and replication network. Table 2-1 describes these networks.

NOTE

When deploying the HyperReplication feature, configure IPv4 addresses instead of IPv6
addresses for the following networks.

Table 2-1 Network descriptions and requirements

Networ Description Requirement

k Type

Manag Used for system management and At least one GE network port is
ement maintenance. needed.
networ
k

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 8

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Networ Description Requirement

k Type

Service Used for communication between Port bonding is recommended

networ compute nodes and the Virtual to aggregate network ports
k Block Service (VBS) by using the between compute nodes and
iSCSI protocol. storage nodes. This will improve
network redundancy reliability.
NIC and network types:
● TCP/IP: Ports can be bonded
in bond1 or bond4 mode.
You are advised to configure
the bond4 mode across NICs
(only for NICs of the same
model). At least two 10GE or
25GE ports are planned.
● IB: Ports can be bonded only
in bond1 mode. In the IB
network, the bond1 mode
enables load balancing. You
are advised to configure the
bond mode across NICs
(only for NICs of the same
model). At least two 56
Gbit/s or 100 Gbit/s ports
are planned.
● RoCE: It is recommended
that ports be bonded in
bond1 mode. Ports cannot
be bonded across NICs. If the
Hi1822 NIC is used, ports 0
and 1 of the NIC are bonded
or ports 2 and 3 are bonded
by default. For details about
other binding combinations,
see A.1 Modifying the RoCE
Configuration File. At least
two 25GE ports are required.
● Local link IPv6 addresses
starting with FE80 and
multicast IPv6 addresses
starting with FF00 cannot be
used. HyperMetro and
HyperReplication are not
supported when IPv6
addresses are used.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 9

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Networ Description Requirement

k Type

Storage Used for data communication Port bonding is recommended

networ between VBS and Object Storage to aggregate network ports
k Device (OSD) processes, between between compute nodes and
VBS and REP processes, and storage nodes. This will improve
between OSD processes. network redundancy reliability.
NIC and network types:
● TCP/IP: Ports can be bonded
in bond1 or bond4 mode.
You are advised to configure
the bond4 mode across NICs
(only for NICs of the same
model). At least two 10GE or
25GE ports are planned.
● IB: Ports can be bonded only
in bond1 mode. In the IB
network, the bond1 mode
enables load balancing. You
are advised to configure the
bond mode across NICs
(only for NICs of the same
model). At least two 56
Gbit/s or 100 Gbit/s ports
are planned.
● RoCE: It is recommended
that ports be bonded in
bond1 mode. Ports cannot
be bonded across NICs. If the
Hi1822 NIC is used, ports 0
and 1 of the NIC are bonded
or ports 2 and 3 are bonded
by default. For details about
other binding combinations,
see A.1 Modifying the RoCE
Configuration File. At least
two 25GE ports are required.
● Local link IPv6 addresses
starting with FE80 and
multicast IPv6 addresses
starting with FF00 cannot be
used. HyperMetro and
HyperReplication are not
supported when IPv6
addresses are used.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 10

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Networ Description Requirement

k Type

Replicat Used for data transmission ● Supports IP networks (GE,

ion between primary and secondary 10GE, and 25GE).
networ storage systems. ● It is recommended that the
k distance between a primary
storage system and a
secondary storage system be
less than 3000 km.
● The bidirectional connection
bandwidth is greater than or
equal to 10 Mbit/s, and the
average write bandwidth of
replication volumes is no
greater than the remote
replication bandwidth.
● Ensure that each replication
node uses two physical
network ports configured
with dual IP addresses for
load sharing. Do not set the
two ports to Bond mode.
● Replication links are fully
interconnected, that is, the
replication network IP
address of each node can
communicate with that of
each node in the remote
storage system at the same
network segment.
● If the replication network IP
addresses of the primary site
belong to different network
segments and the IP
addresses in different
network segments need to
communicate with each
other, the communication
between sites must pass
through the Layer 3
network. Otherwise, the
HyperReplication feature
may be unavailable after the
replication network is
intermittently disconnected.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 11

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

2.1.2 Deployment Solutions

The replication service can be deployed on storage nodes or independent
replication nodes.

● If the replication service is deployed on storage nodes, the replication service

and storage service are deployed in converged mode.
● If the replication service is deployed on independent replication nodes, the
replication service and storage service are deployed independently.

If the number of storage nodes is less than or equal to 16, converged deployment
is recommended. If the number of storage nodes is greater than 16, independent
deployment is recommended.

If the independent deployment mode is used, you are advised not to deploy all
nodes of the replication cluster in the same cabinet and not to deploy control
nodes of the replication cluster in the same cabinet, preventing the entire
replication cluster from becoming faulty due to cabinet power-off.

NOTE

Management nodes can be deployed on VMs, external physical servers, or storage nodes.
In the networking schemes provided in this section, the management nodes are deployed
on external physical servers.

The following describes the six deployment solutions of the HyperReplication

feature.

Solution 1
Figure 2-1 and Figure 2-2 show the deployment scheme and physical networking,
respectively. In this solution:

Basic networking: Compute nodes and storage nodes are deployed independently,
VBS is deployed on storage nodes, and front-end and back-end storage networks
are converged.

Replication network: The replication service and storage service are converged.

Figure 2-1 Deployment scheme

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 12

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Figure 2-2 Physical networking

Table 2-2 lists the network requirements.

Table 2-2 Network requirements

Device Type Requirement IP Address Planning

Compute It is recommended that two Configure one service IP

node network ports form a port address (that is, iSCSI IP
bond to connect to the service address).
network.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 13

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Management ● Provides one management The system has two

node network port to connect to management nodes running in
the management network active/standby mode.
when the management ● If the management
network is not isolated. network is not isolated:
● Provides one management Configure one
network port and one DeviceManager
internal management management IP address for
network port to connect to cluster management.
the management network Configure one
when the management management IP address for
network is isolated. each management node.
● If the management
network is isolated:
Configure one
DeviceManager
management IP address for
cluster management.
Configure one
management IP address for
each management node.
Configure one node
management IP address for
the internal management
network.
Configure one
management IP address for
the internal management
network on each
management node.

Storage node It is recommended that two Configure one management IP

network ports form a port address for each storage node.
bond to connect to the
management network.

It is recommended that two Configure one service IP

network ports form a port address (iSCSI IP address) for
bond to connect to the service each storage node.
network.

Provides two network ports to Configure one storage IP

form a port bond for access to address for each storage node.
the storage network.

Provides one BMC network Configure one BMC network

port to connect to the BMC/ IP address for each storage
management network. node.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 14

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Storage node Provides two network ports Configure two replication IP

running the (not bonded) to connect to addresses for each node
replication the replication network. running the replication service.
service

BMC/ Provides one management Configure one switch

Management network port to connect to management IP address for
switch the management network. remote maintenance.

Storage ● Multiple switches form an Configure one switch

switch M-LAG (recommended) or management IP address for
a stack. remote maintenance.
● In the RoCE network, NOTE
VLANs and PFC flow Two stacked switches function as
one logical switch and therefore
control need to be
require only one switch
configured. management IP address, floating
● If the storage network port on the ETH management
is set to bond4, you need to network ports of the two
switches.
configure trunks.
● Configure the IP address of
the management network
port on a switch.

Replication Provides one management Configure one switch

switch network port to connect to management IP address for
the management network. remote maintenance.

Service switch Switch stack is recommended. Service switches are provided

by customers and their IP
addresses are planned by the
customers.

Solution 2
Figure 2-3 and Figure 2-4 show the deployment scheme and physical networking,
respectively. In this solution:
Basic networking: Compute nodes and storage nodes are deployed independently,
VBS is deployed on storage nodes, and front-end and back-end storage networks
are converged.
Replication network: The replication service and storage service are deployed
independently.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 15

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Figure 2-3 Deployment scheme

Figure 2-4 Physical networking

Table 2-3 lists the network requirements.

Table 2-3 Network requirements

Device Type Requirement IP Address Planning

Compute It is recommended that two Configure one service IP

node network ports form a port address (that is, iSCSI IP
bond to connect to the service address).
network.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 16

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Management ● Provides one management When the system has two

node network port to connect to management nodes running in
the management network active/standby mode
when the management (recommended):
network is not isolated. ● If the management
● Provides one management network is not isolated:
network port and one Configure one
internal management DeviceManager
network port to connect to management IP address for
the management network cluster management.
when the management Configure one
network is isolated. management IP address for
each management node.
● If the management
network is isolated:
Configure one
DeviceManager
management IP address for
cluster management.
Configure one
management IP address for
each management node.
Configure one node
management IP address for
the internal management
network.
Configure one
management IP address for
the internal management
network on each
management node.
When the system has only one
management node:
● If the management
network is not isolated:
Configure one
management IP address for
the management node.
● If the management
network is isolated:
Configure one
management IP address for
the management node.
Configure one
management IP address for
the internal management
network.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 17

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Storage node It is recommended that two Configure one management IP

network ports form a port address for each storage node.
bond to connect to the
management network.

It is recommended that two Configure one service IP

network ports form a port address (iSCSI IP address) for
bond to connect to the service each storage node.
network.

Provides two network ports to Configure one storage IP

form a port bond for access to address for each storage node.
the storage network.

Provides one BMC network Configure one BMC network IP

port to connect to the BMC/ address for each storage node.
management network.

Replication It is recommended that two Configure one management IP

node network ports form a port address for each replication
bond to connect to the node.
management network.

Provides two network ports to Configure one storage IP

form a port bond for access to address for each replication
the storage network. node.

Provides two network ports Configure two replication IP

(not bonded) to connect to addresses for each replication
the replication network. node.

Provides one BMC network Configure one BMC network IP

port to connect to the BMC/ address for each replication
management network. node.

BMC/ Provides one management Configure one switch

Management network port to connect to management IP address for
switch the management network. remote maintenance.

Storage ● Multiple switches form an Configure one switch

switch M-LAG (recommended) or management IP address for
a stack. remote maintenance.
● In the RoCE network, NOTE
VLANs and PFC flow Two stacked switches function as
one logical switch and therefore
control need to be
require only one switch
configured. management IP address, floating
● If the storage network port on the ETH management
is set to bond4, you need to network ports of the two
switches.
configure trunks.
● Configure the IP address of
the management network
port on a switch.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 18

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Replication Provides one management Configure one switch

switch network port to connect to management IP address for
the management network. remote maintenance.

Service switch Switch stack is recommended. Service switches are provided

by customers and their IP
addresses are planned by the
customers.

Solution 3
Figure 2-5 and Figure 2-6 show the deployment scheme and physical networking,
respectively. In this solution:
Basic networking: Compute nodes and storage nodes are deployed independently,
VBS is deployed on compute nodes, and front-end and back-end storage networks
are converged.
Replication network: The replication service and storage service are converged.

Figure 2-5 Deployment scheme

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 19

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Figure 2-6 Physical networking

Table 2-4 lists the network requirements.

Table 2-4 Network requirements

Device Type Requirement IP Address Planning

Compute Provides one network port to Configure one management IP

node connect to the management address.
network.

It is recommended that two Configure one service IP

network ports form a port address (that is, iSCSI IP
bond and connect to the address).
storage network (the service
and storage networks share
physical cables).

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 20

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

CVM Provides three virtual network Configure one management IP

(required only ports: address.
when ● Virtual network port 1: Configure one storage IP
connecting to functions as a management address.
VMware) network port and Configure one service IP
corresponds to the network address (that is, iSCSI IP
port connecting to the address).
management network on
the compute node.
● Virtual network port 2:
functions as a storage
network port and
corresponds to the network
port connecting to the
storage network on the
compute node.
● Virtual network port 3:
functions as a service
network port and
corresponds to the network
port connecting to the
storage network on the
compute node.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 21

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Management ● Provides one management When the system has two

node network port to connect to management nodes running in
the management network active/standby mode
when the management (recommended):
network is not isolated. ● If the management
● Provides one management network is not isolated:
network port and one Configure one
internal management DeviceManager
network port to connect to management IP address for
the management network cluster management.
when the management Configure one
network is isolated. management IP address for
each management node.
● If the management
network is isolated:
Configure one
DeviceManager
management IP address for
cluster management.
Configure one
management IP address for
each management node.
Configure one node
management IP address for
the internal management
network.
Configure one
management IP address for
the internal management
network on each
management node.
When the system has only one
management node:
● If the management
network is not isolated:
Configure one
management IP address for
the management node.
● If the management
network is isolated:
Configure one
management IP address for
the management node.
Configure one
management IP address for
the internal management
network.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 22

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Storage node It is recommended that two Configure one management IP

network ports form a port address for each storage node.
bond to connect to the
management network.

Provides two network ports to Configure one storage IP

form a port bond for access to address for each storage node.
the storage network.

Provides one BMC network Configure one BMC network

port to connect to the BMC/ IP address for each storage
management network. node.

Storage node Provides two network ports Configure two replication IP

running the (not bonded) to connect to addresses for each node
replication the replication network. running the replication service.
service

BMC/ Provides one management Configure one switch

Management network port to connect to management IP address for
switch the management network. remote maintenance.

Storage ● Multiple switches form an Configure one switch

switch M-LAG (recommended) or management IP address for
a stack. remote maintenance.
● In the RoCE network, NOTE
VLANs and PFC flow Two stacked switches function as
one logical switch and therefore
control need to be
require only one switch
configured. management IP address, floating
● If the storage network port on the ETH management
is set to bond4, you need to network ports of the two
switches.
configure trunks.
● Configure the IP address of
the management network
port on a switch.

Replication Provides one management Configure one switch

switch network port to connect to management IP address for
the management network. remote maintenance.

Solution 4
Figure 2-7 and Figure 2-8 show the deployment scheme and physical networking,
respectively. In this solution:
Basic networking: Compute nodes and storage nodes are deployed independently,
VBS is deployed on compute nodes, and front-end and back-end storage networks
are converged.
Replication network: The replication service and storage service are deployed
independently.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 23

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Figure 2-7 Deployment scheme

Figure 2-8 Physical networking

Table 2-5 lists the network requirements.

Table 2-5 Network requirements

Device Type Requirement IP Address Planning

Compute Provides one network port to Configure one management IP

node connect to the management address.
network.

It is recommended that two Configure one service IP

network ports form a port address (that is, iSCSI IP
bond and connect to the address).
storage network (the service
and storage networks share
physical cables).

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 24

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

CVM Provides three virtual network Configure one management IP

(required only ports: address.
when ● Virtual network port 1: Configure one storage IP
connecting to functions as a management address.
VMware) network port and Configure one service IP
corresponds to the network address (that is, iSCSI IP
port connecting to the address).
management network on
the compute node.
● Virtual network port 2:
functions as a storage
network port and
corresponds to the network
port connecting to the
storage network on the
compute node.
● Virtual network port 3:
functions as a service
network port and
corresponds to the network
port connecting to the
storage network on the
compute node.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 25

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Management ● Provides one management When the system has two

node network port to connect to management nodes running in
the management network active/standby mode:
when the management ● If the management
network is not isolated. network is not isolated:
● Provides one management Configure one
network port and one DeviceManager
internal management management IP address for
network port to connect to cluster management.
the management network Configure one
when the management management IP address for
network is isolated. each management node.
● If the management
network is isolated:
Configure one
DeviceManager
management IP address for
cluster management.
Configure one
management IP address for
each management node.
Configure one node
management IP address for
the internal management
network.
Configure one
management IP address for
the internal management
network on each
management node.

Storage node It is recommended that two Configure one management IP

network ports form a port address for each storage node.
bond to connect to the
management network.

Provides two network ports to Configure one storage IP

form a port bond for access to address for each storage node.
the storage network.

Provides one BMC network Configure one BMC network

port to connect to the BMC/ IP address for each storage
management network. node.

Replication It is recommended that two Configure one management IP

node network ports form a port address for each replication
bond to connect to the node.
management network.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 26

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Provides two network ports to Configure one storage IP

form a port bond for access to address for each replication
the storage network. node.

Provides two network ports Configure two replication IP

(not bonded) to connect to addresses for each replication
the replication network. node.

Provides one BMC network Configure one BMC network

port to connect to the BMC/ IP address for each replication
management network. node.

BMC/ Provides one management Configure one switch

Management network port to connect to management IP address for
switch the management network. remote maintenance.

Storage ● Multiple switches form an Configure one switch

switch M-LAG (recommended) or management IP address for
a stack. remote maintenance.
● In the RoCE network, NOTE
VLANs and PFC flow Two stacked switches function as
one logical switch and therefore
control need to be
require only one switch
configured. management IP address, floating
● If the storage network port on the ETH management
is set to bond4, you need to network ports of the two
switches.
configure trunks.
● Configure the IP address of
the management network
port on a switch.

Replication Provides one management Configure one switch

switch network port to connect to management IP address for
the management network. remote maintenance.

Solution 5
Figure 2-9 and Figure 2-10 show the deployment scheme and physical
networking, respectively. In this solution:
Basic networking: Compute nodes and storage nodes are deployed independently,
VBS is deployed on compute nodes, and front-end and back-end storage networks
are converged.
Replication network: The replication service and storage service are converged.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 27

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Figure 2-9 Deployment scheme

Figure 2-10 Physical networking

Table 2-6 lists the network requirements.

Table 2-6 Network requirements

Device Type Requirement IP Address Planning

Compute Provides one network port to Configure one management IP

node connect to the management address.
network.

It is recommended that two Configure one storage IP

network ports form a port address.
bond and connect to the
storage network (the service
and storage networks share
physical cables).

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 28

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Management ● Provides one management When the system has two

node network port to connect to management nodes running in
the management network active/standby mode:
when the management ● If the management
network is not isolated. network is not isolated:
● Provides one management Configure one
network port and one DeviceManager
internal management management IP address for
network port to connect to cluster management.
the management network Configure one
when the management management IP address for
network is isolated. each management node.
● If the management
network is isolated:
Configure one
DeviceManager
management IP address for
cluster management.
Configure one
management IP address for
each management node.
Configure one node
management IP address for
the internal management
network.
Configure one
management IP address for
the internal management
network on each
management node.

Storage node It is recommended that two Configure one management IP

network ports form a port address for each storage node.
bond to connect to the
management network.

Provides two network ports to Configure one storage IP

form a port bond for access to address for each storage node.
the storage network.

Provides one BMC network Configure one BMC network

port to connect to the BMC/ IP address for each storage
management network. node.

Storage node Provides two network ports Configure two replication IP

running the (not bonded) to connect to addresses for each node
replication the replication network. running the replication service.
service

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 29

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

BMC/ Provides one management Configure one switch

Management network port to connect to management IP address for
switch the management network. remote maintenance.

Storage ● Multiple switches form an Configure one switch

switch M-LAG (recommended) or management IP address for
a stack. remote maintenance.
● In the RoCE network, NOTE
VLANs and PFC flow Two stacked switches function as
one logical switch and therefore
control need to be
require only one switch
configured. management IP address, floating
● If the storage network port on the ETH management
is set to bond4, you need to network ports of the two
switches.
configure trunks.
● Configure the IP address of
the management network
port on a switch.

Replication Provides one management Configure one switch

switch network port to connect to management IP address for
the management network. remote maintenance.

Solution 6
Figure 2-11 and Figure 2-12 show the deployment scheme and physical
networking, respectively. In this solution:

Basic networking: Compute nodes and storage nodes are deployed independently,
VBS is deployed on compute nodes, and front-end and back-end storage networks
are converged.

Replication network: The replication service and storage service are deployed
independently.

Figure 2-11 Deployment scheme

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 30

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Figure 2-12 Physical networking

Table 2-7 lists the network requirements.

Table 2-7 Network requirements

Device Type Requirement IP Address Planning

Compute Provides one network port to Configure one management IP

node connect to the management address.
network.

It is recommended that two Configure one service IP

network ports form a port address.
bond and connect to the
storage network (the service
and storage networks share
physical cables).

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 31

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Management ● Provides one management When the system has two

node network port to connect to management nodes running in
the management network active/standby mode:
when the management ● If the management
network is not isolated. network is not isolated:
● Provides one management Configure one
network port and one DeviceManager
internal management management IP address for
network port to connect to cluster management.
the management network Configure one
when the management management IP address for
network is isolated. each management node.
● If the management
network is isolated:
Configure one
DeviceManager
management IP address for
cluster management.
Configure one
management IP address for
each management node.
Configure one node
management IP address for
the internal management
network.
Configure one
management IP address for
the internal management
network on each
management node.

Storage node It is recommended that two Configure one management IP

network ports form a port address for each storage node.
bond to connect to the
management network.

Provides two network ports to Configure one storage IP

form a port bond for access to address for each storage node.
the storage network.

Provides one BMC network Configure one BMC network

port to connect to the BMC/ IP address for each storage
management network. node.

Replication It is recommended that two Configure one management IP

node network ports form a port address for each replication
bond to connect to the node.
management network.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 32

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Device Type Requirement IP Address Planning

Provides two network ports to Configure one storage IP

form a port bond for access to address for each replication
the storage network. node.

Provides two network ports Configure two replication IP

(not bonded) to connect to addresses for each replication
the replication network. node.

Provides one BMC network Configure one BMC network

port to connect to the BMC/ IP address for each replication
management network. node.

BMC/ Provides one management Configure one switch

Management network port to connect to management IP address for
switch the management network. remote maintenance.

Storage ● Multiple switches form an Configure one switch

switch M-LAG (recommended) or management IP address for
a stack. remote maintenance.
● In the RoCE network, NOTE
VLANs and PFC flow Two stacked switches function as
one logical switch and therefore
control need to be
require only one switch
configured. management IP address, floating
● If the storage network port on the ETH management
is set to bond4, you need to network ports of the two
switches.
configure trunks.
● Configure the IP address of
the management network
port on a switch.

Replication Provides one management Configure one switch

switch network port to connect to management IP address for
the management network. remote maintenance.

2.2 System Requirements

Before deploying HyperReplication, ensure that the system meets the
requirements listed in Table 2-8.

Table 2-8 System requirements

Item Requirement

Compatibilit For details about the operating systems supported by servers,

y access Huawei Storage Interoperability Navigator.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 33

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Item Requirement

Reserved Space is reserved in primary and secondary storage pools for

space snapshots.
Formula: Reserved space > Average host write bandwidth x
Synchronization period (unit: second)

CPU and ● Replication and storage nodes deployed on the same servers:
memory Each node requires 8 GB extra memory to deploy the
resources replication service.
occupied by ● Replication nodes deployed on independent servers:
the Each node requires 20 vCPUs and 64 GB memory.
replication
software

Metadata Space used by CCDB: Two partitions are required, each with at
space least 10 GB capacity. It is recommended that each partition have
25 GB capacity.

Number of A replication cluster can contain three, five, seven, or nine control
control nodes:
nodes in a ● If the number of nodes in the replication service cluster is less
replication than 27, you are advised to select three control nodes to form
cluster the replication cluster.
● If the number of nodes in the replication service cluster is
greater than or equal to 27 but less than 45, you are advised
to select five control nodes to form the replication cluster.
● If the number of nodes in the replication service cluster is
greater than or equal to 45 but less than or equal to 64, you
are advised to select seven control nodes to form the
replication cluster.
● A replication control cluster can be configured with a
maximum of nine nodes, and a maximum of four faulty nodes
are tolerated in the cluster, improving the cluster reliability.

2.3 Required Software Packages and Documentation

Before deploying HyperReplication, obtain the software packages and
documentation listed in Table 2-9.

Table 2-9 Required software packages and documentation

Item Description How to Obtain

Software Storage software See Installation Preparations of the

packages installation package desired scenario in the FusionStorage
and other dependency Block Storage Software Installation
packages Guide.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 34

FusionStorage
Block Storage HyperReplication Feature Guide 2 Planning and Preparation

Item Description How to Obtain

FusionStora ● Provides guidance For enterprise users, log in to https://

ge Block for obtaining basic support.huawei.com/enterprise. For
Storage software carrier users, log in to https://
Software installation support.huawei.com. In the search box,
Installation packages and enter FusionStorage and click the
Guide other dependency associated path that is displayed. The
packages. FusionStorage product page is displayed.
● Provides guidance Search for and download required
for software documents.
installation.

FusionStora Provides guidance for

ge Block planning and
Storage deploying networks.
Network
Planning
Guide

FusionStora Provides guidance for

ge Block configuring basic
Storage services.
Basic
Service
Configurati
on Guide

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 35

FusionStorage
Block Storage HyperReplication Feature Guide 3 Software Installation

3 Software Installation

To ensure that subsequent operations can be performed properly, configure

switches and iBMC IP addresses, prepare management VMs, and install CVMs,
operating systems for servers, and storage software by following instructions in
FusionStorage Block Storage Software Installation Guide.

NOTE

● If the replication service and storage service are deployed on different nodes, you do not
need to install replication nodes during software installation. Select the replication
nodes when creating a replication cluster.
● If the replication service and storage service are deployed on same nodes, select a
desired number of storage nodes to run the replication service when creating a
replication cluster.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 36

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

4 Configuring HyperReplication

4.1 Configuration Process

4.2 Checking the License
4.3 Configuring Basic Services
4.4 Creating a Replication Cluster
Creating a replication cluster includes creating a replication control cluster and a
replication service cluster, configuring the replication network, and authenticating
the replication cluster. You must create a replication cluster in both primary and
secondary storage systems.
4.5 Adding a Remote Device
This operation establishes a logical connection between the primary and
secondary storage systems for data transmission.
4.6 Creating a Remote Replication Pair
Creating a replication pair between the local and remote directories enables data
in the local directory to be backed up to the remote cluster. When a disaster
occurs, you can use the remote backup data to reconstruct data on the local end.
Alternatively, you can directly take over services from the remote end to minimize
data loss and shorten service interruption time.
4.7 Creating a Remote Replication Consistency Group
In medium- and large-sized database applications, data, logs, and change records
are stored in associated volumes of storage systems. If data of one volume is lost,
data in other volumes becomes invalid. To ensure data availability, you can add
remote replication pairs of these volumes to the same consistency group.

4.1 Configuration Process

This section describes how to configure HyperReplication.
Figure 4-1 shows the process for configuring HyperReplication.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 37

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Figure 4-1 Process for configuring HyperReplication

4.2 Checking the License

A license file is a credential for using the HyperReplication feature. Before
configuring HyperReplication, ensure that its license file is valid.

Context
You must check the license on both the primary and secondary storage systems.
The HyperReplication feature requires a standard license or an advanced license.
Table 4-1 shows their application scenarios.

Table 4-1 Application scenarios of HyperReplication licenses

Standard Edition Advanced Edition

RPO ≥ 1 minute RPO ≥ 10 seconds

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 38

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Procedure
Step 1 Log in to DeviceManager.
Step 2 Choose Settings > License Management.
The current license information is displayed.

----End

Follow-Up Procedure
If no HyperReplication license is available, apply for and import a license file. For
details about how to apply for and import a license file, see the license operation
guide of the corresponding product model.

4.3 Configuring Basic Services

Configure basic services on both primary and secondary storage systems, including
creating storage pools, VBS, volumes, hosts, and host groups, and connecting to
compute nodes.
For details, see Basic Service Configuration Guide.

NOTE

It is recommended that the configurations of storage pools and volumes in primary and
secondary storage systems be the same.
The sizes of volumes in the primary and secondary storage systems must be the same.

4.4 Creating a Replication Cluster

Creating a replication cluster includes creating a replication control cluster and a
replication service cluster, configuring the replication network, and authenticating
the replication cluster. You must create a replication cluster in both primary and
secondary storage systems.

Context
A replication cluster consists of a replication control cluster and a replication
service cluster:
● A replication control cluster manages cluster nodes and metadata and
contains three, five, seven, or nine nodes. The CCDB process of the replication
control cluster requires local storage space to record metadata information.
● A replication service cluster manages remote replication pairs and consistency
groups and contains 3 to 64 nodes.

Prerequisites
The IP address of the port that connects each replication cluster node to the
replication network has been obtained. To obtain the IP address, perform the
following operations:

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 39

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

1. Log in to a replication cluster node using its management IP address.

2. Run the ifconfig command to query port information.
[root@HN_0_0 ~]# ifconfig
eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.27.11 netmask 255.255.192.0 broadcast 192.168.63.255
inet6 fe80::f816:3eff:fe06:409 prefixlen 64 scopeid 0x20<link>
ether XX:XX:XX:XX:XX:XX txqueuelen 1000 (Ethernet)
RX packets 137182 bytes 647468977 (617.4 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 90603 bytes 15569803 (14.8 MiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

eth1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500

inet 192.168.78.18 netmask 255.255.192.0 broadcast 192.168.127.255
inet6 fe80::f816:3eff:fecc:8504 prefixlen 64 scopeid 0x20<link>
ether XX:XX:XX:XX:XX:XX txqueuelen 1000 (Ethernet)
RX packets 14385213 bytes 3319786835 (3.0 GiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 12994523 bytes 3536755766 (3.2 GiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

eth2: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500

inet 192.168.164.16 netmask 255.255.192.0 broadcast 192.168.191.255
inet6 fe80::f816:3eff:fea6:8a93 prefixlen 64 scopeid 0x20<link>
ether XX:XX:XX:XX:XX:XX txqueuelen 1000 (Ethernet)
RX packets 12270 bytes 1441625 (1.3 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 12151 bytes 1033156 (1008.9 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

In the command output, inet is the IP address of each port.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Replication Cluster.

Step 3 Click Create.

The Create Replication Cluster page is displayed.

Step 4 Create a replication control cluster.

● Recommended

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 40

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

a. Set the name of the replication control cluster.

NOTE

▪ A replication control cluster name contains 1 to 64 characters.

▪ A replication control cluster name contains only letters, digits, underscores

(_), and hyphens (-).
b. In the Selected Nodes area, view node configurations.
c. Click Submit.
d. Wait until the creation is successful, and then click Next.
● Custom

a. Set the name of the replication control cluster.

NOTE

▪ A replication control cluster name contains 1 to 64 characters.

▪ A replication control cluster name contains only letters, digits, underscores

(_), and hyphens (-).
b. In Metadata Storage Location, select the location to store the metadata.
Table 4-2 describes the deployment of replication service metadata disks
in a scenario where remote replication and HyperMetro services are
enabled and replication nodes and storage nodes are deployed together.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 41

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Table 4-2 Metadata disk selection for replication services in converged

deployment scenarios
Basic Service Replication Service Metadata Disk Selection
Metadata Disk
(ZooKeeper Disk)
Type

Independent SAS The replication service metadata disk and the

disks or SSDs basic service ZooKeeper disk are deployed
together.

Independent SATA Select an independent SAS disk or SSD as the

disks replication service metadata disk, or select a
system disk partition as the replication service
metadata disk.

System disk (SSD) Select a system disk partition as the replication

partition service metadata disk.

System disk (non- Select an independent SAS disk or SSD as the

SSD) partition replication service metadata disk.

NOTE

▪ The size of the metadata disk must be greater than 105 GB.

▪ If a system disk partition is used as the replication service metadata disk, the
system disk must meet the following requirements:
○ If the number of pools is less than or equal to 4, the system disk is a SAS
disk or SSD.
○ If the number of pools is greater than 4, the system disk is an SSD.

▪ If you select Physical Disk, specify the disk type and slot selection
mode.
NOTE

The slot selection mode can be:

○ Specify Slot: Manually enter a slot number. The value ranges from 0 to
35.
○ Manually Select: Manually select disks for each node.

▪ If you select System Disk Partition, the system stores metadata in

the /opt/ccdb_disk2 partition of the system disks.
c. Select the nodes to create the replication control cluster.

▪ If Metadata Storage Location is set to Physical Disk, select the

nodes and disks for creating the replication control cluster from
Available Nodes.

▪ If Metadata Storage Location is set to Partition, select the nodes

for creating the replication control cluster from Available Nodes.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 42

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

d. Click Submit.
e. Wait until the creation is successful, and then click Next.
Step 5 Create a replication service cluster.
1. Select the nodes to create the replication service cluster.
2. Click Submit.
3. Wait until the creation is successful, and then click Next.
Step 6 Configure the replication network.
NOTE

The system supports two methods for configuring the replication network. Method 2 is
recommended.
● Method 1: Modify the network port configuration file to configure the replication
network in advance (for details, see A.2 Configuring the Replication Network by
Modifying the Configuration File), and then read the existing storage network IP
address by following the instructions in this section.
● Method 2: Configure the replication network by following the instructions provided in
this section.

1. Set the parameters listed in Table 4-3.

Table 4-3 Replication network parameters

Parameter Description

Transmission Protocol Indicates the transmission protocol

of the replication network. TCP is
supported.

IP Address Indicates the start IP address of the

IP address segment to which the
replication IP addresses belong.

Subnet Mask Indicates the subnet mask of the

replication network.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 43

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Parameter Description

Port Indicates the port used by an IP

address.
NOTE
If the ports of replication IP addresses
have been configured in the
configuration file in advance, skip this
parameter.

2. Click Preview.
The system will query the replication IP address assigned to each replication
node.
3. Click Submit.
4. Click Next.

NOTE

Log in to a replication node and run the ifconfig command. If the replication node does not

have an IP address for the replication network, click next to the replication node and
choose Modify to configure an IP address.
In the scenario of interconnecting with FusionCompute, you are not allowed to modify the

configuration information by clicking next to the replication node because the network
configuration of the remote replication port has been completed on FusionCompute.

Step 7 Authenticate the replication cluster.

Table 4-4 describes the related parameters.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 44

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Table 4-4 Cluster authentication parameters

Parameter Description

Pre-Shared Key Label Identifies the pre-shared key of the

cluster.
[Value range]
● A pre-shared key label contains 5 to
32 characters.
● A pre-shared key label contains
only letters, digits, and underscores
(_), and starts with a letter.

Pre-Shared Key Implements identity authentication

together with Pre-Shared Key Label.
[Value range]
● A pre-shared key contains 8 to 31
characters.
● A pre-shared key contains at least
one special character and at least
two of the following character
types: uppercase letters, lowercase
letters, and digits. Special
characters include !"#$%&'()*
+,-./:;<=>?@[\]^`{_|}~ and spaces.

Confirm Pre-Shared Key Confirms the pre-shared key.

Step 8 Click Submit.

Step 9 Click Finish.

----End

Follow-Up Procedure
If different replication ports on a replication node use IP addresses on the same
network segment, you must configure policy-based routes for these ports.
Otherwise, these ports cannot be accessed.
1. Log in to the replication node using its management IP address.
2. Configure route tables.
NOTE

The value range of the route priority is from 0 to 255. Check the priorities of existing
routes written in the /etc/iproute2/rt_tables file, and ensure that the priority values
of their routing tables do not conflict.
For example, run the following two commands to configure route table
rep_eth0 with priority 200 for eth0 and configure route table rep_eth1 with
priority 201 for eth1. Configure this parameter as required.
[root@localhost ~]# echo "200 rep_eth0" >> /etc/iproute2/rt_tables
[root@localhost ~]# echo "201 rep_eth1" >> /etc/iproute2/rt_tables

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 45

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

3. Configure policy-based routing rules based on the routing tables.

Table 4-5 provides an example of the replication port configuration. The
configuration should be adjusted based on actual conditions.

Table 4-5 Replication port configuration example

Networ Routing IP Address Network Gateway

k Port Table Name Segment/
Name Subnet Mask

eth0 rep_eth0 192.168.10.31 192.168.10.0/2 192.168.10.1

4

eth1 rep_eth1 192.168.10.32 192.168.10.0/2 192.168.10.1

4

[root@localhost ~]# ip route flush table rep_eth0

[root@localhost ~]# ip route flush table rep_eth1
[root@localhost ~]# ip rule add from 192.168.10.31 table rep_eth0
[root@localhost ~]# ip rule add from 192.168.10.32 table rep_eth1
[root@localhost ~]# ip route add 192.168.10.0/24 dev eth0 src 192.168.10.31 table rep_eth0
[root@localhost ~]# ip route add 192.168.10.0/24 dev eth1 src 192.168.10.32 table rep_eth1
[root@localhost ~]# ip route add default dev eth0 via 192.168.10.1 table rep_eth0
[root@localhost ~]# ip route add default dev eth1 via 192.168.10.1 table rep_eth1

4. Verify the configuration results of policy-based routing.

a. Query the configured policy-based routing information.
[root@localhost ~]# ip route list table rep_eth0
default via 192.168.10.31 dev eth0
192.168.10.0/24 dev eth0 scope link src 192.168.10.31

[root@localhost ~]# ip route list table rep_eth1

default via 192.168.10.32 dev eth1
192.168.10.0/24 dev eth1 scope link src 192.168.10.32

[root@localhost ~]# ip rule list

0: from all lookup local
32764: from 192.168.10.32 lookup rep_eth1
32765: from 192.168.10.31 lookup rep_eth0
32766: from all lookup main
32767: from all lookup default

b. After the configuration is complete, run the ping command to check

whether each network port can communicate with the target replication
plane.
[root@localhost ~]# ping -I 192.168.10.31 XXX.XXX.XXX.XXX
[root@localhost ~]# ping -I 192.168.10.32 XXX.XXX.XXX.XXX

NOTE

In this command, 192.168.10.31 indicates the replication IP address configured

for the replication plane, and XXX.XXX.XXX.XXX indicates the IP address of the
destination replication plane.

If the data returned by the replication plane can be received, the policy-based
routing configuration has taken effect. Otherwise, check whether the default
gateway configuration of the management plane and the policy-based
routing configuration of the replication plane are correct.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 46

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

5. Write the policy-based routing configuration to the configuration file.

The routing information configured in 3 is temporary. To prevent
configuration loss after a restart, perform the following steps to write the
routing configuration to the configuration file.
The following example describes how to configure the routing information of
eth0 in various operating systems:
NOTE

In the following operations, 192.168.10.31 indicates the replication IP address

configured for eth0, and 192.168.10.1 indicates the gateway configured for eth0.
– For Red Hat, Oracle, CentOS, and distributed storage OS:
i. Run the vi /etc/sysconfig/network-scripts/route-eth0 command to
edit the configuration file. Enter the following information and save
the file:
192.168.10.0/24 dev eth0 src 192.168.10.31 table rep_eth0
default dev eth0 via 192.168.10.1 table rep_eth0

ii. Run the vi /etc/sysconfig/network-scripts/rule-eth0 command to

edit the configuration file. Enter the following information and save
the file:
from 192.168.10.31 table rep_eth0

iii. Repeat 5.i and 5.ii to write the policy-based routing configurations of
other network adapters to the configuration file.
– For SUSE:
i. Run the vi /etc/sysconfig/network/ifroute-eth0 command to edit
the configuration file. Enter the following information and save the
file:
ip route flush table rep_eth0
ip route add 192.168.10.0/24 dev eth0 src 192.168.10.31 table rep_eth0
ip route add default dev eth0 via 192.168.10.1 table rep_eth0
ip rule del table rep_eth0
ip rule add from 192.168.10.31 table rep_eth0

ii. Modify the permission on the policy-based routing file and write it to
the eth0 network configuration file.
[root@localhost ~]# chmod +x /etc/sysconfig/network/ifroute-eth0
[root@localhost ~]# echo "POST_UP_SCRIPT=\`/etc/sysconfig/network/ifroute-eth0\`"
>> /etc/sysconfig/network/ifcfg-eth0

iii. Repeat 5.i and 5.ii to write the policy-based routing configurations of
other network adapters to the configuration file.

4.5 Adding a Remote Device

This operation establishes a logical connection between the primary and
secondary storage systems for data transmission.

Prerequisites
The local and remote storage systems communicate properly.

Procedure
Step 1 Log in to DeviceManager.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 47

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Step 2 Choose Data Protection > Replication Cluster.

Step 3 Select the cluster for which you want to add a remote device, click , and select
Add Remote Device.

The Add Remote Device dialog box is displayed.

Step 4 Set the remote device parameters listed in Table 4-6.

Table 4-6 Remote device parameters

Parameter Description

Remote Device Replication IP Address Indicates the replication IP address of

a device in the remote replication
cluster.
NOTE
If there are multiple devices in the remote
replication cluster, you can use any of
them.

Pre-Shared Key Label Identifies the pre-shared key of the

cluster.
[Value range]
● A pre-shared key label contains 5 to
32 characters.
● A pre-shared key label contains
only letters, digits, and underscores
(_), and starts with a letter.
The value must be the same as the
value of Pre-Shared Key Label
specified in 4.4 Creating a Replication
Cluster.

Pre-Shared Key Implements identity authentication

together with Pre-Shared Key Label.
The value must be the same as the
value of Pre-Shared Key specified in
4.4 Creating a Replication Cluster.

Step 5 Click OK.

----End

4.6 Creating a Remote Replication Pair

Creating a replication pair between the local and remote directories enables data
in the local directory to be backed up to the remote cluster. When a disaster
occurs, you can use the remote backup data to reconstruct data on the local end.
Alternatively, you can directly take over services from the remote end to minimize
data loss and shorten service interruption time.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 48

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Prerequisites
● The remote replication license is valid.
● The local and remote storage systems support the remote replication
function.
● The remote volume of the remote replication pair is not mapped to the host.
● The local volume is not configured with features that are mutually exclusive
with the remote replication function.
● The remote resource ID has been obtained. Obtain the remote resource ID:
Log in to the remote device, choose Services > Block Service > Volume, click
, select the desired resource ID, and check the ID of the remote resource
whose capacity is the same as the local resource capacity.

Procedure
Step 1 Log in to DeviceManager.
Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.
Step 3 Click Create.
The Create Remote Replication Pair dialog box is displayed.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 49

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Step 4 Configure basic information about the remote replication pair. Table 4-7 describes
related parameters.

Table 4-7 Remote replication pair parameters

Parameter Description

Synchronization Method Indicates the data synchronization mode between

primary and secondary resources. Possible options
are as follows:
● Manual: Data must be manually synchronized
between primary and secondary resources.
● Timed wait after synchronization begins: The
system starts timing as soon as the latest
synchronization begins. When the time specified
in Interval has elapsed, the system
automatically starts to synchronize data between
primary and secondary resources again.
NOTE
If the latest synchronization is still being executed
after the time specified in Interval has elapsed, data
synchronization will start after the latest
synchronization finishes.
● Timed wait after synchronization ends: The
system starts timing as soon as the latest
synchronization ends. When the time specified in
Interval has elapsed, the system automatically
starts to synchronize data between primary and
secondary resources again.
● Specific time policy: The system automatically
synchronizes data between primary and
secondary resources based on a specific time
policy.

Interval Indicates the waiting time for the two

synchronization methods, Timed wait after
synchronization begins and Timed wait after
synchronization ends.
[Value range]
● An integer ranging from 1 to 1440, when
minutes is used as the unit.
● An integer ranging from 10 to 86,400, when
seconds is used as the unit.
NOTE
This parameter is available only when Synchronization
Method is set to Timed wait after synchronization
begins or Timed wait after synchronization ends.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 50

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Parameter Description

Time Policy Specific time policy: The system automatically

synchronizes data between primary and secondary
resources based on a specific time policy.
[Value range]
● Daily: Data synchronization begins on every day.
Its value ranges from 00:00 to 23:59.
● Hourly: Data synchronization begins in every
hour. Its value is an integer ranging from 0 to 59.
NOTE
This parameter is available only when Synchronization
Method is set to Specific time policy.

Initial Synchronization Indicates the initial data synchronization mode

between the primary resource and secondary
resource. Possible options are as follows:
● Data is inconsistent between the primary and
secondary resources. After the remote
replication pair is created, the system
automatically performs data synchronization.:
After a remote replication task is created, data
on the secondary resource is overwritten
immediately. Ensure that the data on the
secondary resource is no longer necessary or has
been backed up.
● The data on primary and secondary resources
is consistent and data synchronization is not
required.: Ensure that the data on the primary
and secondary resources is consistent. If the data
is inconsistent, the secondary resource cannot be
used to recover data or take over services from
the primary resource when the primary resource
fails. As a result, data loss and service
interruption may occur. This option is not
displayed when the resource type is File System.

Remote Replication Indicates the consistency group to which the newly

Consistency Group created remote replication pair belongs.
NOTE
If no remote replication consistency group is available, you
can click Create to create a remote replication consistency
group and add the remote replication pair to the
consistency group.

Step 5 Select the local and remote resources of the remote replication pair.
1. Select a replication cluster in Local Device and Remote Device respectively.
2. In the Local Resource pane, select the desired local resource and enter the
remote resource ID.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 51

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

NOTE

Log in to the remote device, choose Services > Block Service > Volume, click ,
select the desired resource ID, and check the ID of the remote resource.
3. Click Synchronization Parameters and set advanced properties of the remote
replication pair. Table 4-8 describes related parameters. Then click OK.

Table 4-8 Remote replication synchronization parameters

Parameter Description

Speed Indicates the data synchronization

speed of the remote data
replication. The value ranges from 1
MB/s to 1024 MB/s (or 1024 KB/s to
1,048,576 KB/s). The default value is
32 MB/s.

Recovery Policy Indicates whether the system

automatically synchronizes data
after a remote replication fault is
rectified. The default value is
Automatic.
[Value range]
– Automatic: The previous ongoing
tasks automatically resume
immediately after the remote
replication fault is rectified.
– Manual: The value of Running
Status changes to To be
recovered after the remote
replication fault is rectified. In
this case, you can check the
system and then determine
whether to resume the previous
ongoing tasks manually.

Step 6 Click OK.

----End

4.7 Creating a Remote Replication Consistency Group

In medium- and large-sized database applications, data, logs, and change records
are stored in associated volumes of storage systems. If data of one volume is lost,
data in other volumes becomes invalid. To ensure data availability, you can add
remote replication pairs of these volumes to the same consistency group.

Prerequisites
● Role of the remote replication pair to be added to a consistency group is
Primary.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 52

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

● The secondary resources in remote replication pairs to be added to a

consistency group must be from the same storage device.
● The replication mode of the remote replication pair to be added to the
consistency group must be the same as that of the consistency group.

Precautions
If a remote replication pair is not in split status, the system will split the remote
replication pair first and then add it to a consistency group.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Click Create.

The Create Remote Replication Consistency Group page is displayed.

Step 4 Set the basic information about the consistency group. Table 4-9 describes related
parameters.

Table 4-9 Consistency group parameters

Parameter Description

Name Indicates the name of a remote replication consistency group.

[Value range]
● A name must be unique.
● A name can only contain digits, letters, underscores (_),
and hyphens (-) and must start with a digit or letter.
● A name must contain 1 to 31 characters.

Replication Indicates a replication cluster to which the consistency group

Cluster belongs.

Remote Device Indicates a remote storage device to which the secondary

resource of the remote replication pair to be added to the
consistency group belongs.

Step 5 Optional: Add a remote replication pair to the remote replication consistency
group.
1. Select Add Remote Replication Pair.
2. In the Available Remote Replication Pairs pane, select one or more remote
replication pairs and click to add the selected objects to the Selected
Remote Replication Pairs pane.

Step 6 Click Advanced to set advanced properties of the remote replication consistency
group. Table 4-10 describes related parameters.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 53

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Table 4-10 Advanced parameters of a remote replication consistency group

Parameter Description

Speed Indicates the data synchronization

speed of a remote replication
consistency group. The value ranges
from 1 MB/s to 1024 MB/s (or 1024
KB/s to 1,048,576 KB/s). The default
value is 32 MB/s.

Recovery Policy Indicates whether the system

automatically synchronizes data after
a remote replication fault is rectified.
The default value is Automatic.
[Value range]
● Automatic: The previous ongoing
tasks automatically resume
immediately after the remote
replication fault is rectified.
● Manual: The value of Running
Status changes to To be recovered
after the remote replication fault is
rectified. In this case, you can check
the system and then determine
whether to resume the previous
ongoing tasks manually.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 54

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Parameter Description

Synchronization Method Indicates the data synchronization

mode between primary and secondary
resources.
[Value range]
● Manual: Data must be manually
synchronized between primary and
secondary resources.
● Timed wait after synchronization
begins: The system starts timing as
soon as the latest synchronization
begins. When the time specified in
Interval has elapsed, the system
automatically starts to synchronize
data between primary and
secondary resources again.
NOTE
If the latest synchronization is still
being executed after the time specified
in Interval has elapsed, data
synchronization will start after the
latest synchronization finishes.
● Timed wait after synchronization
ends: The system starts timing as
soon as the latest synchronization
ends. When the time specified in
Interval has elapsed, the system
automatically starts to synchronize
data between primary and
secondary resources again.
● Specific time policy: The system
automatically synchronizes data
between primary and secondary
resources based on a specific time
policy.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 55

FusionStorage
Block Storage HyperReplication Feature Guide 4 Configuring HyperReplication

Parameter Description

Interval Indicates the waiting time for the two

synchronization methods, Timed wait
after synchronization begins and
Timed wait after synchronization
ends.
[Value range]
● An integer ranging from 1 to 1440,
when minutes is used as the unit.
● An integer ranging from 10 to
86,400, when seconds is used as
the unit.
NOTE
This parameter is available only when
Synchronization Method is set to Timed
wait after synchronization begins or
Timed wait after synchronization ends.

Time Policy The system automatically synchronizes

data between primary and secondary
resources based on a specific time
policy.
[Value range]
● Daily: Data synchronization begins
on every day. Its value ranges from
00:00 to 23:59.
● Hourly: Data synchronization
begins in every hour. Its value is an
integer ranging from 0 to 59.
NOTE
This parameter is available only when
Synchronization Method is set to Specific
time policy.

Step 7 Click OK.

----End

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 56

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

5 Managing HyperReplication

5.1 Managing a Remote Replication Pair

This section describes operations related to remote replication management. The
operations include viewing, synchronizing, splitting, modifying, and deleting a
remote replication pair, enabling/disabling write protection of secondary resources,
and performing a primary/secondary switchover.
5.2 Managing a Remote Replication Consistency Group
After creating a remote replication consistency group, you may need to manage it
due to service adjustment, DR, data testing, or new data security requirements.
5.3 Managing a Replication Cluster

5.1 Managing a Remote Replication Pair

This section describes operations related to remote replication management. The
operations include viewing, synchronizing, splitting, modifying, and deleting a
remote replication pair, enabling/disabling write protection of secondary resources,
and performing a primary/secondary switchover.

5.1.1 Viewing Information About a Remote Replication Pair

This operation enables you to view information about existing remote replication
pairs and to monitor their status.

Procedure
Step 1 Log in to DeviceManager.
Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.
Step 3 View information about an existing remote replication pair. Table 5-1 describes
related parameters.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 57

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Table 5-1 Remote replication pair parameters

Parameter Description

Local Resource Indicates the name of a local resource.

Remote Resource Indicates the name of a remote

resource.

Pair ID Indicates the ID of a remote

replication pair.

Role Indicates whether the local resource of

the remote replication pair is the
primary resource. If local resources are
primary resources, it displays as
Primary. Otherwise, it displays as
Secondary.

Health Status Indicates whether the remote

replication pair is functioning normally.
Possible options are as follows:
● Normal: The pair is functioning
normally. You can perform the
synchronization operation.
● Fault: The primary resource or
secondary resource of the pair
malfunctions or the replication link
is down. Therefore, you cannot
perform the synchronization
operation.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 58

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Parameter Description

Running Status Indicates whether the remote

replication pair is running properly.
Possible options are as follows:
● Normal: indicates that the data
synchronization of the pair is
completed successfully.
● Synchronizing: indicates that the
data of the pair is being
synchronized.
● Split: indicates that the pair is in
the split status, and the
synchronization is suspended.
● Interrupted: indicates that internal
I/Os of the remote replication is
faulty or the replication link is
down. Therefore, you cannot
perform the synchronization
operation temporarily.
● To be recovered: indicates that
after the faulty internal I/Os of the
remote replication are rectified or
the replication link is up, and if
Recovery Policy is Manual, the pair
is to be recovered. You need to
manually start data synchronization
for the pair.
● Invalid: indicates that the pair is
unusable and cannot be restored in
this status, except deleting it. You
must delete the pair on both
primary and secondary storage
devices. Otherwise, the pair creation
may fail.

Remote Replication Consistency Group Indicates the ID of a remote

ID replication consistency group.

Added to Remote Replication Indicates whether to add a remote

Consistency Group replication pair to a remote replication
consistency group.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 59

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Parameter Description

Replication Mode Indicates the data replication mode.

Asynchronous: After data is written to
primary resources successfully, the
host receives a write success response.
Meanwhile, the primary site records
differences between the written data
and the secondary resource data. Then
the changed data recorded by the
primary site is synchronized to the
secondary site, minimizing the write
latency for upper-layer applications.

Secondary Resource Data Status Indicates the status of the secondary

resource. Possible options are as
follows:
● Consistent: Data on the secondary
resource is a copy of the data on
the primary resource (at the time
when the latest successful
synchronization starts). At this time,
data on the secondary resource is
available but not necessarily the
same as the current data on the
primary resource.
● Inconsistent: Data on the
secondary resource is not a copy of
the data on the primary resource
(at the current point in time or at
the time when the previous
synchronization is ended). In this
state, data on the secondary
resource is unavailable. You can
synchronize the data to ensure data
consistency between the primary
resource and secondary resource.

Secondary Resource Protection Indicates the protection status of the

secondary resource. Once the
protection for the secondary resource
in a remote replication pair is enabled,
the application server cannot modify
the data on the secondary resource.
Therefore, you can ensure the security
of the data on the secondary resource.

Step 4 3. Optional: Click to view the local resource name, remote resource name,
recovery policy, speed, start and end time of synchronization, and synchronization
progress of each remote replication pair.
----End

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 60

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

5.1.2 Modifying Properties of a Remote Replication Pair

This operation enables you to change the synchronization type, speed, and
recovery policy of a remote replication pair.

Precautions
The properties of a remote replication pair which has been added to a consistency
group cannot be modified.

Procedure
Step 1 Log in to DeviceManager.
Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.

Step 3 Select the desired remote replication pair, click on the right, and select Modify.
The Modify page is displayed on the right.
Step 4 Modify properties of the remote replication pair. Table 5-2 describes related
parameters.

Table 5-2 Remote replication pair parameters

Parameter Description

Speed Indicates the data synchronization speed of the

remote data replication. The value ranges from 1
MB/s to 1024 MB/s (or 1024 KB/s to 1,048,576
KB/s).

Recovery Policy Indicates whether the system automatically

synchronizes data after a remote replication fault is
rectified.
[Value range]
● Automatic: The previous ongoing tasks
automatically resume immediately after the
remote replication fault is rectified.
● Manual: The value of Running Status changes
to To be recovered after the remote replication
fault is rectified. In this case, you can check the
system and then determine whether to resume
the previous ongoing tasks manually.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 61

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Parameter Description

Synchronization Method Indicates the data synchronization mode between

primary and secondary resources. Possible options
are as follows:
● Manual: Data must be manually synchronized
between primary and secondary resources.
● Timed wait after synchronization begins: The
system starts timing as soon as the latest
synchronization begins. When the time specified
in Interval has elapsed, the system
automatically starts to synchronize data between
primary and secondary resources again.
NOTE
If the latest synchronization is still being executed
after the time specified in Interval has elapsed, data
synchronization will start after the latest
synchronization finishes.
● Timed wait after synchronization ends: The
system starts timing as soon as the latest
synchronization ends. When the time specified in
Interval has elapsed, the system automatically
starts to synchronize data between primary and
secondary resources again.
● Specific time policy: The system automatically
synchronizes data between primary and
secondary resources based on a specific time
policy.

Interval Indicates the waiting time for the two

synchronization methods, Timed wait after
synchronization begins and Timed wait after
synchronization ends.
[Value range]
● An integer ranging from 1 to 1440, when
minutes is used as the unit.
● An integer ranging from 10 to 86,400, when
seconds is used as the unit.
NOTE
This parameter is available only when Synchronization
Method is set to Timed wait after synchronization
begins or Timed wait after synchronization ends.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 62

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Parameter Description

Time Policy Specific time policy: The system automatically

synchronizes data between primary and secondary
resources based on a specific time policy.
[Value range]
● Daily: Data synchronization begins on every day.
Its value ranges from 00:00 to 23:59.
● Hourly: Data synchronization begins in every
hour. Its value is an integer ranging from 0 to 59.
NOTE
This parameter is available only when Synchronization
Method is set to Specific time policy.

Step 5 Click OK.

----End

5.1.3 Synchronizing a Remote Replication Pair

After the initial synchronization of a remote replication pair, if an application
server writes data to primary resources or you need to split the pair, data
inconsistency between primary and secondary resources may occur. This operation
enables you to synchronize the remote replication pair to ensure data consistency
between primary and secondary resources.

Prerequisites
● The remote replication pair for which you want to perform data
synchronization has not been added to a consistency group.
● Secondary resource protection has been enabled.
● Running Status of the remote replication pair supports synchronization.
Table 5-3 shows in which status a remote replication pair can be
synchronized.

Table 5-3 Status requirements for synchronizing a remote replication pair

Running Status of a Remote Synchronization Supported or Not

Replication Pair

Normal √

Split √

Interrupted (link down) ×

Interrupted (link up) √

To be recovered √

Synchronizing ×

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 63

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Running Status of a Remote Synchronization Supported or Not

Replication Pair

Invalid ×

√: Supported
×: Not supported

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.

Step 3 Select the remote replication pair that you want to synchronize and click
Synchronize.
NOTE

You can also click on the right of the remote replication pair and select Synchronize.

Step 4 Confirm your operation as prompted.

----End

5.1.4 Splitting a Remote Replication Pair

This operation enables you to split a remote replication task if you want to enable
secondary resources to save the data of primary resources at a specific point in
time, interrupt the remote replication link for maintenance, or expand the network
bandwidth. After a remote replication pair is split, the data written to primary
resources by the application server will not be synchronized to the secondary
resource. This may cause data inconsistency between primary and secondary
resources.

Prerequisites
● The remote replication pair that you want to split has not been added to a
consistency group.
● Running Status of the remote replication pair supports splitting. Table 5-4
shows in which status a remote replication pair can be split.

Table 5-4 Status requirements for splitting a remote replication pair

Running Status of a Remote Splitting Supported or Not

Replication Pair

Normal √

Split ×

Interrupted (link down) √ (can be performed only on the

primary storage system.)

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 64

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Running Status of a Remote Splitting Supported or Not

Replication Pair

Interrupted (link up) √

To be recovered √

Synchronizing √

Invalid ×

√: Supported
×: Not supported

Precautions
If the initial synchronization of a remote replication pair is not completed, splitting
the remote replication pair may cause the unavailability of secondary resources.
Therefore, exercise caution when performing this operation.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.

Step 3 Select the remote replication pair that you want to split and click Split.
NOTE

You can also click on the right of the remote replication pair and select Split.

Step 4 Confirm your operation as prompted.

----End

5.1.5 Performing a Primary/Secondary Switchover

This operation enables you to switch the roles of the primary and secondary
resources of a remote replication pair. The data synchronization direction changes
as the primary/secondary relationship changes.

Prerequisites
● The remote replication pair on which you want to perform a primary/
secondary switchover has not been added to a consistency group
● A primary/secondary switchover can be performed on an asynchronous
remote replication pair when the following requirements are met:
Running Status of the pair is set to Split, Secondary Resource Data Status
is set to Consistent, and Secondary Resource Protection is set to Disable.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 65

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.

Step 3 Select the desired remote replication pair, click , and select Primary/Secondary
Switchover.

Step 4 Confirm your operation as prompted.

----End

5.1.6 Enabling Protection for Secondary Resources

This operation enables you to ensure that only the data on primary resources in a
remote replication pair can be written to secondary resources.

Prerequisites
Running Status and Secondary Resource Data Status of the remote replication
pair support protection enablement. Table 5-5 shows in which status protection
for secondary resources in a remote replication pair can be enabled.

Table 5-5 Status requirements for enabling protection for secondary resources in a
remote replication pair

Running Status of a Secondary Resource Protection Enablement

Remote Replication Data Status Supported or Not
Pair

Normal Any status ×

Split Consistent √

Interrupted (link down) Consistent √ (can be performed

only on the secondary
storage system.)

Interrupted (link up) Any status ×

To be recovered Any status ×

Synchronizing Any status ×

Invalid Any status ×

√: Supported
×: Not supported

Procedure
Step 1 Log in to DeviceManager.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 66

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.

Step 3 Select the desired remote replication pair, click on the right, and select Enable
Protection for Secondary Resource.

----End

5.1.7 Disabling Protection for Secondary Resources

When the remote device needs to take over services, disable protection for
secondary resources to make resources readable and writable. Therefore, write
data services can be provided for upper-layer services. At this time, roles of the
primary and secondary directories in a remote replication pair stay unchanged but
primary resources cannot synchronize data to secondary resources.

Prerequisites
Running Status and Secondary Resource Data Status of the remote replication
pair support protection disablement. Table 5-6 shows in which status protection
for secondary resources in a remote replication pair can be disabled.

Table 5-6 Status requirements for disabling protection for secondary resources in
a remote replication pair

Running Status of a Secondary Resource Protection Disablement

Remote Replication Data Status Supported or Not
Pair

Normal Any status ×

Split Consistent √

Interrupted (link down) Consistent √ (can be performed

only on the secondary
storage system.)

Interrupted (link up) Any status ×

To be recovered Any status ×

Synchronizing Any status ×

Invalid Any status ×

√: Supported
×: Not supported

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 67

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Step 3 Select the desired remote replication pair, click on the right, and select Disable
Protection for Secondary Resource.

----End

5.1.8 Deleting a Remote Replication Pair

After you delete a remote replication pair, the replication relationship between
primary and secondary resources of the remote replication pair will be deleted,
and data on primary resources will no longer be synchronized to secondary
resources.

Prerequisites
● The remote replication pair to be deleted has not been added to a consistency
group.
● Running Status of the remote replication pair supports deletion. Table 5-7
shows in which status a remote replication pair can be deleted.

Table 5-7 Status requirements for deleting a remote replication pair

Running Status of a Remote Deletion Supported or Not

Replication Pair

Normal ×

Split √

Interrupted √

To be recovered ×

Synchronizing ×

Invalid √

√: Supported
×: Not supported

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication Pair.

Step 3 Select the desired remote replication pair and click Delete.
NOTE

You can also click on the right of the remote replication pair and select Delete.

Step 4 Confirm your operation as prompted.

----End

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 68

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

5.2 Managing a Remote Replication Consistency Group

After creating a remote replication consistency group, you may need to manage it
due to service adjustment, DR, data testing, or new data security requirements.

5.2.1 Viewing Remote Replication Consistency Group

Information
This operation enables you to view information about an existing remote
replication consistency group, so that you can effectively monitor the status of the
consistency group.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 View the basic information of an existing remote replication consistency group.
Table 5-8 describes related parameters.

Table 5-8 Remote replication consistency group parameters

Parameter Description

Name Indicates the name of a remote

replication consistency group.

Health Status Indicates whether a remote replication

consistency group is functioning
properly. Possible options are as
follows:
● Normal: The consistency group is
functioning properly. You can
perform the synchronization
operation.
● Fault: The consistency group
malfunctions or the replication link
is down, you cannot perform the
synchronization operation.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 69

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Parameter Description

Running Status Indicates whether a remote replication

consistency group is running properly.
Possible options are as follows:
● Normal: Data has been
synchronized in the consistency
group.
● Synchronizing: Data is being
synchronized in the consistency
group.
● Split: The consistency group is in
split status, and data
synchronization is suspended.
● Interrupted: An internal I/O error
occurs or the replication link is
down, and you cannot perform the
synchronization operation
temporarily.
● To be recovered: After the internal
I/O error is rectified or the
replication link is up, the
consistency group enters this status
if Recovery Policy is Manual. You
need to manually start
synchronization for the consistency
group to synchronize data.
● Invalid: The consistency group is
unusable and cannot be restored in
this status, except deleting it. You
must delete the consistency group
on both primary and secondary
storage devices. Otherwise, you
may fail to create any after the
deletion.

Replication Mode Indicates the data replication mode of

a remote replication consistency group.
Asynchronous: After data is written to
primary resources successfully, the
host receives a write success response.
Meanwhile, the primary site records
differences between the written data
and the secondary resource data. Then
the changed data recorded by the
primary site is synchronized to the
secondary site, minimizing the write
latency for upper-layer applications.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 70

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Parameter Description

Role Indicates whether local resources of a

remote replication pair in a
consistency group are primary
resources. If local resources are
primary resources, it displays as
Primary. Otherwise, it displays as
Secondary.

Recovery Policy Indicates whether the system

automatically synchronizes data after
a remote replication fault is rectified.
The default value is Automatic.
[Value range]
● Automatic: The previous ongoing
tasks automatically resume
immediately after the remote
replication fault is rectified.
● Manual: The value of Running
Status changes to To be recovered
after the remote replication fault is
rectified. In this case, you can check
the system and then determine
whether to resume the previous
ongoing tasks manually.

Speed Indicates the data synchronization

speed of a remote replication
consistency group. The value ranges
from 1 MB/s to 1024 MB/s (or 1024
KB/s to 1,048,576 KB/s). The default
value is 32 MB/s.

Time Difference Indicates the synchronization time

difference between primary and
secondary storage systems.
● When Running Status is Normal,
this parameter indicates the time
difference between the current time
and the start time of the current
replication.
● When Running Status is
Synchronizing, this parameter
indicates the time difference
between the current time and the
start time of the last replication.

----End

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 71

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

5.2.2 Modifying Properties of a Remote Replication

Consistency Group
This operation enables you to modify the rate, recovery policy, or synchronization
type of a remote replication consistency group.

Precautions
After the properties of a remote replication consistency group are modified, the
properties of pairs in the consistency group remain unchanged. However, the
working properties are consistent with those of the consistency group.

Procedure
Step 1 Log in to DeviceManager.
Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Select the remote replication consistency group you want to modify, click on the
right, and select Modify.
The Modify page is displayed on the right.
Step 4 Modify properties of the remote replication consistency group. Table 5-9 describes
related parameters.

Table 5-9 Remote replication consistency group parameters

Parameter Description

Speed Indicates the data synchronization speed of a remote

replication consistency group.
The value ranges from 1 MB/s to 1024 MB/s (or 1024 KB/s to
1,048,576 KB/s).

Recovery Indicates whether the system automatically synchronizes data

Policy after a remote replication fault is rectified.
[Value range]
● Automatic: The previous ongoing tasks automatically
resume immediately after the remote replication fault is
rectified.
● Manual: The value of Running Status changes to To be
recovered after the remote replication fault is rectified. In
this case, you can check the system and then determine
whether to resume the previous ongoing tasks manually.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 72

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Parameter Description

Synchronizatio Indicates the data synchronization mode between primary and

n Method secondary resources. Possible options are as follows:
● Manual: Data must be manually synchronized between
primary and secondary resources.
● Timed wait after synchronization begins: The system
starts timing as soon as the latest synchronization begins.
When the time specified in Interval has elapsed, the system
automatically starts to synchronize data between primary
and secondary resources again.
NOTE
If the latest synchronization is still being executed after the time
specified in Interval has elapsed, data synchronization will start
after the latest synchronization finishes.
● Timed wait after synchronization ends: The system starts
timing as soon as the latest synchronization ends. When the
time specified in Interval has elapsed, the system
automatically starts to synchronize data between primary
and secondary resources again.
● Specific time policy: The system automatically synchronizes
data between primary and secondary resources based on a
specific time policy.

Interval Indicates the waiting time for the two synchronization

methods, Timed wait after synchronization begins and
Timed wait after synchronization ends.
[Value range]
● An integer ranging from 1 to 1440, when minutes is used as
the unit.
● An integer ranging from 10 to 86,400, when seconds is used
as the unit.
NOTE
This parameter is available only when Synchronization Method is set
to Timed wait after synchronization begins or Timed wait after
synchronization ends.

Time Policy The system automatically synchronizes data between primary

and secondary resources based on a specific time policy.
[Value range]
● Daily: Data synchronization begins on every day. Its value
ranges from 00:00 to 23:59.
● Hourly: Data synchronization begins in every hour. Its value
is an integer ranging from 0 to 59.
NOTE
This parameter is available only when Synchronization Method is set
to Specific time policy.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 73

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Step 5 Click OK.

----End

5.2.3 Synchronizing a Remote Replication Consistency Group

This operation enables you to synchronize all remote replication pairs in a
consistency group. After the synchronization, the local data and remote data in
each remote replication pair are consistent at the synchronization time point,
ensuring data validity.

Prerequisites
● This operation can be performed only on the primary device of a remote
replication consistency group.
● Secondary resource protection has been enabled.
● Running Status of the remote replication consistency group supports
synchronization. Table 5-10 shows in which status a remote replication
consistency group can be synchronized.

Table 5-10 Status requirements for synchronizing a remote replication

consistency group

Running Status of a Remote Synchronization Supported or Not

Replication Consistency Group

Normal √

Split √

Interrupted (link down) ×

Interrupted (link up) √ (can be performed only on the

primary device of a consistency
group.)

To be recovered √

Invalid ×

Synchronizing ×

√: Supported
×: Not supported

Procedure
Step 1 Log in to DeviceManager.
Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.
Step 3 Select the remote replication consistency group you want to synchronize and click
Synchronize.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 74

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

NOTE

You can also click on the right of the remote replication consistency group and select
Synchronize.

Step 4 Confirm your operation as prompted.

----End

5.2.4 Splitting a Remote Replication Consistency Group

This operation enables you to split all remote replication pairs in a consistency
group at the same time. A complete point-in-time duplicate is generated for
primary resources of the remote replication pairs, and can be used for report
generation, testing, data mining, decision-making, and data recovery.

Precautions
Running Status of the remote replication consistency group supports splitting.
Table 5-11 shows in which status remote replication pairs in a consistency group
can be split.

Table 5-11 Status requirements for splitting remote replication pairs in a

consistency group

Running Status of a Remote Splitting Supported or Not

Replication Consistency Group

Normal √

Split ×

Interrupted (link down) √ (can be performed only on the

primary device of a consistency group.)

Interrupted (link up) √

To be recovered √

Invalid ×

Synchronizing √

√: Supported
×: Not supported

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Select the remote replication consistency group you want to split and click Split.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 75

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

NOTE

You can also click on the right of the remote replication consistency group and select
Split.

Step 4 Confirm your operation as prompted.

----End

5.2.5 Performing a Primary/Secondary Switchover for a

Remote Replication Consistency Group
This operation enables you to switch the roles of primary and secondary resources
in a remote replication consistency group. If primary resources in the remote
replication consistency group are faulty, you can perform a primary/secondary
switchover. This operation enables secondary resources to become new primary
resources, replacing the faulty primary resources to ensure service data continuity.

Prerequisites
Running Status of the remote replication consistency group and Secondary
Resource Data Status of remote replication pairs in the consistency group
support a primary/secondary switchover. Table 5-12 shows in which status a
primary/secondary switchover can be performed for a remote replication
consistency group.

Table 5-12 Status requirements for performing a primary/secondary switchover

for a remote replication consistency group
Running Status of a Secondary Resource Switchover Supported
Remote Replication Data Status or Not
Consistency Group

Normal Consistent ×

Normal Inconsistent ×

Split Consistent √

Split Inconsistent ×

Interrupted Consistent ×

Interrupted Inconsistent ×

To be recovered Consistent ×

To be recovered Inconsistent ×

Invalid Any status ×

Synchronizing Any status ×

√: Supported
×: Not supported

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 76

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Select the desired remote replication consistency group, click on the right, and
select Primary/Secondary Switchover.

Step 4 Confirm your operation as prompted.

----End

5.2.6 Enabling Protection for Secondary Resources in a

Remote Replication Consistency Group
This operation enables you to enable protection for secondary resources of all
remote replication pairs in a consistency group. After this operation, data cannot
be written to secondary resources, ensuring secondary resource data security.

Prerequisites
Running Status of the remote replication consistency group and Secondary
Resource Data Status of remote replication pairs in the consistency group
support protection enablement. Table 5-13 shows in which status protection for
secondary resources in a remote replication consistency group can be enabled.

Table 5-13 Status requirements for enabling protection for secondary resources in
a remote replication consistency group

Running Status of a Secondary Resource Protection Enablement

Remote Replication Data Status Supported or Not
Consistency Group

Normal Any status ×

Split Consistent √

Interrupted (link down) Consistent √ (can be performed

only on the secondary
storage system.)

Interrupted (link up) Any status ×

To be recovered Any status ×

Synchronizing Any status ×

Invalid Any status ×

√: Supported
×: Not supported

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 77

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Select the desired remote replication consistency group, click on the right, and
select Enable Protection for Secondary Resource.

----End

5.2.7 Disabling Protection for Secondary Resources in a

Remote Replication Consistency Group
This operation enables you to disable protection for secondary resources of all
remote replication pairs in a consistency group. After this operation, data can be
written to secondary resources.

Prerequisites
Running Status of the remote replication consistency group and Secondary
Resource Data Status of remote replication pairs in the consistency group
support protection disablement. Table 5-14 shows in which status protection for
secondary resources in a remote replication consistency group can be disabled.

Table 5-14 Status requirements for disabling protection for secondary resources in
a remote replication consistency group

Running Status of a Secondary Resource Protection Disablement

Remote Replication Data Status Supported or Not
Consistency Group

Normal Any status ×

Split Consistent √

Interrupted (link down) Consistent √ (can be performed

only on the secondary
storage system.)

Interrupted (link up) Any status ×

To be recovered Any status ×

Synchronizing Any status ×

Invalid Any status ×

√: Supported
×: Not supported

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 78

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Select the desired remote replication consistency group, click on the right, and
select Disable Protection for Secondary Resource.

----End

5.2.8 Adding a Remote Replication Pair

This operation enables you to add a remote replication pair into a consistency
group. This ensures that the secondary resource data of the new pair and that of
other pairs in the consistency group are consistent in time.

Prerequisites
● Replication Mode of the remote replication pair to be added to a consistency
group is the same as that of the consistency group.
● Role of the consistency group to which the remote replication pair is to be
added is Primary.
● Role of the remote replication pair to be added to a consistency group is
Primary.
● The number of remote replication pairs in a consistency group has not
reached the upper limit.

Precautions
● If a consistency group is in split status, when a remote replication pair is
added, the system will split the remote replication pair first and then add it
into the consistency group.
● If a consistency group is not in split status, when a remote replication pair is
added, the system will split the consistency group first, then split the remote
replication pair, add it into the group, and synchronize the consistency group.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Select the desired remote replication consistency group, click on the right, and
select Add Remote Replication Pair.

The Add Remote Replication Pair page is displayed on the right.

Step 4 In the pair list on the left, select one or more remote replication pairs and click
to add the selected pairs to the pair list on the right.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 79

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Step 5 Click OK.

----End

5.2.9 Removing a Remote Replication Pair

This operation enables you to remove a remote replication pair from a consistency
group. After the remote replication pair is removed, secondary resource data of
the pair and that of other pairs in the consistency group are no longer consistent
in time.

Prerequisites
● The consistency group has been split.
● Role of the consistency group from which the remote replication pair is to be
removed is Primary.

Procedure
Step 1 Log in to DeviceManager.
Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Select the desired remote replication consistency group, click on the right, and
select Remove Remote Replication Pair.
The Remove Remote Replication Pair page is displayed on the right.
Step 4 In the pair list on the left, select one or more remote replication pairs and click
to add the selected pairs to the pair list on the right.
Step 5 Click OK.
Confirm your operation as prompted.

----End

5.2.10 Deleting a Remote Replication Consistency Group

This operation enables you to delete a remote replication consistency group that is
no longer needed.

Prerequisites
Running Status of the remote replication consistency group supports deletion.
Table 5-15 shows in which status a remote replication consistency group can be
deleted.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 80

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Table 5-15 Status requirements for deleting a remote replication consistency

group

Running Status of a Remote Deletion Supported or Not

Replication Consistency Group

Normal ×

Split √

Interrupted √

To be recovered ×

Synchronizing ×

Invalid √

√: Supported
×: Not supported

Precautions
● The essence of deleting a remote replication consistency group is to delete
the consistency group configuration information stored on both local and
remote storage devices.
● If a remote replication link fails, you must forcibly delete the consistency
group from the local and remote devices respectively.
● After a consistency group is deleted, you cannot centrally manage the remote
replication pairs in the group.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Remote Replication > Remote Replication
Consistency Group.

Step 3 Select the remote replication consistency group you want to delete and click
Delete.
NOTE

You can also click on the right of the remote replication consistency group and select
Delete.

Step 4 Confirm your operation as prompted.

----End

5.3 Managing a Replication Cluster

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 81

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

5.3.1 Expanding a Replication Cluster

This operation enables you to expand the capacity of a replication cluster when
the number of nodes in the cluster cannot meet service requirements.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Replication Cluster.

Step 3 Select the desired replication cluster, click on the right, and select Expand
Capacity.

The Expansion page is displayed.

Step 4 Add replication nodes.

1. Select the nodes to be added to the replication cluster.
2. Click Submit.
3. Wait until the submission is successful, and then click Next.

Step 5 Configure the network.

1. Configure network information for replication nodes. Table 5-16 describes
related parameters.

Table 5-16 Replication node network parameters

Parameter Description

Transmission Protocol Indicates the transmission protocol

of the replication network. TCP is
supported.

Start IP Address Indicates the start IP address of the

replication network.

Subnet Mask Indicates the subnet mask. This

parameter identifies the subnet to
which an IP address belongs.

Gateway Indicates the gateway. This

parameter functions together with
Subnet Mask.

Port Indicates the port used by an IP

address.

2. Click Preview.
3. Click Submit.

Step 6 Wait until the submission is successful, and then click Cancel.

----End

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 82

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

5.3.2 Reducing a Replication Cluster

This operation enables you to reduce the capacity of a replication cluster by
removing a replication node from it.

Prerequisites
The number of nodes to be removed from a replication cluster is greater than or
equal to 4.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Replication Cluster.

Step 3 Click on the right of the desired replication cluster and select Reduce Capacity.

The Reduce Capacity page is displayed.

Step 4 Select the nodes to be removed from the replication cluster.

Step 5 Click OK.

----End

5.3.3 Removing a Remote Device

This operation enables you to remove an unnecessary remote device from a
replication cluster.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Replication Cluster.

Step 3 In the Remote Device pane, select the desired remote device, click , and select
Delete.

Step 4 Click OK.

----End

5.3.4 Configuring a Cluster Service Credential

A replication cluster uses cluster authentication credentials for identification and
authentication. To improve system security, you can periodically configure the
cluster authentication credentials for a replication cluster. After configuring a
cluster authentication credential, you need to change the remote cluster
authentication credential of the corresponding remote device in the remote cluster
simultaneously.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 83

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Replication Cluster.

Step 3 Select the desired cluster, click , and select Configure Cluster Service
Credential.

The Configure Cluster Service Credential page is displayed.

Step 4 Set Pre-Shared Key Label, Pre-Shared Key, and Confirm Pre-Shared Key.
NOTE

● Pre-Shared Key Label

– A pre-shared key label contains 5 to 32 characters.
– A pre-shared key label can contain only letters, digits, and underscores (_), and
must start with a letter.
● Pre-Shared Key
– A pre-shared key contains 8 to 31 characters.
– A pre-shared key must contain at least one special character (!"#$%&\'()*
+,-./:;<=>?@[\]^`{_|}~ and space) and at least two of the following character types:
uppercase letters, lowercase letters, and digits.

Step 5 Click OK.

----End

5.3.5 Changing a Replication Cluster Name

This operation enables you to change the name of a replication cluster.

Procedure
Step 1 Log in to DeviceManager.

Step 2 Choose Data Protection > Replication Cluster.

Step 3 Select a replication cluster, click on the right, and select Change Name.

The Change Name page is displayed.

Step 4 Set the replication cluster name.

NOTE

● A replication cluster name contains 1 to 31 characters.

● A replication cluster name can contain only letters, digits, and underscores (_), and must
start with a letter.

Step 5 Click OK.

----End

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 84

FusionStorage
Block Storage HyperReplication Feature Guide 5 Managing HyperReplication

5.3.6 Deleting a Replication Cluster

This operation enables you to delete a replication cluster if the replication services
in it are no longer used.

Procedure
Step 1 Log in to DeviceManager.
Step 2 Choose Data Protection > Replication Cluster.

Step 3 Click on the right of the desired replication cluster and select Delete.
Step 4 Confirm your operation as prompted.

----End

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 85

FusionStorage
Block Storage HyperReplication Feature Guide A Appendix

A Appendix

A.1 Modifying the RoCE Configuration File

When you bond ports over the RoCE protocol, if the Hi1822 NIC is used, ports 0
and 1 of the NIC are bonded or ports 2 and 3 are bonded by default. If other
bonding combinations are required, perform the following steps to modify the
configuration file.

Step 1 Run the following command to open the /etc/modprobe.d/hiroce_pro.conf

configuration file:
vi /etc/modprobe.d/hiroce_pro.conf

The default file content is as follows:

options hiroce g_want_bond_slave_cnt=0x2 g_want_bond0_slave_bits=0x3 g_want_bond1_slave_bits=0xc

NOTE

● g_want_bond_slave_cnt indicates the number of ports to be bonded. To bond two ports

on a 4-port 25GE NIC, set this parameter to 0x2. To bond four ports, set this parameter
to 0x4.
● g_want_bond0_slave_bits indicates the bits of the first group of bonded ports. The bits
correspond to ports 0 to 4 respectively from right to left. Value 1 indicates that a port is
enabled, and value 0 indicates that a port is disabled. For example, 0x3 (0011) indicates
that only ports 0 and 1 are enabled.
● g_want_bond1_slave_bits indicates the bits of the second group of bonded ports. The
bits correspond to ports 0 to 4 respectively from right to left. Value 1 indicates that a
port is enabled, and value 0 indicates that a port is disabled. For example, 0xc (1100)
indicates that only ports 2 and 3 are enabled.
● Table A-1 lists the port combinations.

Table A-1 Port combinations

Bonded Ports Bits Value

Ports 1 and 3 1010 0xa

Ports 0 and 2 0101 0x5

Ports 1 and 2 0110 0x6

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 86

FusionStorage
Block Storage HyperReplication Feature Guide A Appendix

Bonded Ports Bits Value

Ports 0 and 3 1001 0x9

Step 2 For example, to bond ports 1 and 2, change the value of

g_want_bond0_slave_bits to 0x6 (0110).
options hiroce g_want_bond_slave_cnt=0x2 g_want_bond0_slave_bits=0x6 g_want_bond1_slave_bits=0xc

Step 3 After the modification is complete, press Esc to exit the editing mode and
enter :wq! to save the modification and exit.

Step 4 Restart the node where the NIC is located for the modification to take effect.

----End

A.2 Configuring the Replication Network by Modifying

the Configuration File
This section describes how to configure the replication plane network.

Prerequisites
● You have planned two replication plane IP addresses.
● You have obtained the password of user root for logging in to the server.
● You have obtained the management plane IP address of the server.

Procedure
Step 1 Open a browser on the local PC, enter http://BMC IP address of the server in the
address bar, and press Enter to go to the login page.

Step 2 On the login page that is displayed, enter the user name and password to log in
to the BMC system.

The default BMC user name is root and the default password is Huawei@123.

Step 3 Click Remote Control.

NOTE

For some Huawei servers, you need to choose Remote Control > Remote Virtual Console
(sharing mode) to open the remote control window.

The remote control window is displayed.

Step 4 Log in to the server as user root.

Step 5 Run the following commands to go to the path where the configuration file is
saved:
● For Red Hat, Oracle, CentOS, and distributed storage OS, run cd /etc/
sysconfig/network-scripts.
● For SUSE, run the cd /etc/sysconfig/network command.

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 87

FusionStorage
Block Storage HyperReplication Feature Guide A Appendix

Step 6 Use the vi editor to open the configuration file for modifying network port
information. Table A-2 provides an example.
NOTE

The following provides an example of using the vi editor.

1. vi Configuration file name
Example: vi ifcfg-eth0
2. Press i to go to the editing mode.
3. Edit the configuration file.
4. Press Esc and run :wq. Save the change and exit the vi editor.

Table A-2 Network port configuration file information

Name Content (Distributed Storage File Content (SUSE)
OS, Red Hat, Oracle, and
CentOS)

ifcfg-eth0 TYPE=Ethernet BOOTPROTO='static'

PROXY_METHOD=none BROADCAST=''
BROWSER_ONLY=no ETHTOOL_OPTIONS=''
BOOTPROTO=static IPADDR='IP address/Subnet
DEFROUTE=no mask'
IPV4_FAILURE_FATAL=no MTU=''
IPV6INIT=yes NAME='82540EM Gigabit
Ethernet Controller'
IPV6_AUTOCONF=yes
NETWORK=''
IPV6_DEFROUTE=yes
REMOTE_IPADDR=''
IPV6_FAILURE_FATAL=no
STARTMODE='auto'
IPV6_ADDR_GEN_MODE=stabl
e-privacy USERCONTROL='no'
NAME=eth0
UUID=8a06f7f6-
a663-4889-87a9-ddacf4de6e8f
DEVICE=eth0
ONBOOT=yes
NM_CONTROLLED=no
IPADDR=IP address
NETMASK=Subnet mask

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 88

FusionStorage
Block Storage HyperReplication Feature Guide A Appendix

Name Content (Distributed Storage File Content (SUSE)

OS, Red Hat, Oracle, and
CentOS)

ifcfg-eth1 TYPE=Ethernet BOOTPROTO='static'

PROXY_METHOD=none BROADCAST=''
BROWSER_ONLY=no ETHTOOL_OPTIONS=''
BOOTPROTO=static IPADDR='IP address/Subnet
DEFROUTE=no mask'
IPV4_FAILURE_FATAL=no MTU=''
IPV6INIT=yes NAME='82540EM Gigabit
Ethernet Controller'
IPV6_AUTOCONF=yes
NETWORK=''
IPV6_DEFROUTE=yes
REMOTE_IPADDR=''
IPV6_FAILURE_FATAL=no
STARTMODE='auto'
IPV6_ADDR_GEN_MODE=stabl
e-privacy USERCONTROL='no'
NAME=eth1
UUID=5a08g7f6-
a443-4229-77d9-
daacf4de6d6a
DEVICE=eth1
ONBOOT=yes
NM_CONTROLLED=no
IPADDR=IP address
NETMASK=Subnet mask

Step 7 Run the following commands to make the configuration take effect and run the
ifconfig command to check whether the configuration takes effect. If the IP
addresses configured in the preceding step are displayed in the command output,
the configuration takes effect.
ifdown eth0; ifup eth0
ifdown eth1; ifup eth1
ifconfig

Step 8 Configure policy-based routing. For details, see Follow-Up Procedure.

----End

Issue 05 (2021-02-05) Copyright © Huawei Technologies Co., Ltd. 89