Configuration Planning - Huawei OceanStor V3 Series Features Manual

OceanStor V3 Series
Remote Replication Feature Guide for Block
l
l
l
Based on the former analysis, the details of the solution are as follows:
l
l
l

4.1.2 Configuration Planning

This section describes how to plan the remote replication feature for central backup and
disaster recovery. This section is organized into two parts: network planning and service
planning.
Network Planning
Figure 4-2
and disaster recovery is implemented without changing the carrier's existing network.
Issue 06 (2018-01-30)
Table 4-4 Advantages of an asynchronous remote replication solution
Advantage
Incremental data
replication
No host performance
deterioration
Quick service takeover
Data availability
The carrier needs to analyze the backup data and requires that data at site A be quickly
recovered after a disaster. Therefore, data at the backup site must be available anytime.
Service continuity
The backup site must be able to quickly take over services when a disaster causes a
failure of site A to minimize the loss, by setting the Interval to 10 minutes to satisfy the
RPO requirement.
Others
Besides backup and disaster recovery, the carrier hopes backup data be managed
centrally and latest data at service sites be available anytime so that services can be
analyzed conveniently. Therefore, data at service sites needs to be centrally backed up to
one site. In this way, copies of all data can be quickly obtained when necessary.
Choose the central disaster backup solution and build a central backup site at site B
(equipment at the backup site includes an OceanStor storage system, an application
server, a maintenance terminal, and an iSCSI switch).
Use asynchronous remote replication for disaster backup.
Set up an iSCSI connection between sites A and B.
shows a diagram of the planned network. As shown in the diagram, central backup
Huawei Proprietary and Confidential
Copyright © Huawei Technologies Co., Ltd.
Description
Incremental data replication is implemented between the
primary and secondary storage systems. Specifically, after
the initial synchronization (full replication), only data
changes are replicated from the service site to the backup
site during each synchronization process. Since the amount
of data replicated each time is small, synchronization is
fast.
Data replication is between storage systems, occupying no
host resources, so it does not affect current services. The
quality of service (QoS) does not decrease.
When a disaster occurs, remote replication pairs are
interrupted between the service site and the backup site.
The backup site quickly takes over services of the service
site.
4 Configuration Example
75