Updated on 2023-06-16 GMT+08:00

Solution Overview

What Is Two-Site Three-Center?

Two-site three-center is a disaster recovery (DR) solution used for service continuity. With three data center coexisted, the continuity of mission-critical services can be guaranteed even if any two centers fail. This significantly improves the availability of the solution.

The two-site three-center DR solution includes a production center, an intra-city DR center, and a remote DR center.

  • Production center: It provides services for external applications.
  • Intra-city DR center: It is usually dozens of kilometers away from the production center. Applications can be switched to the intra-city DR center without data loss. This is the level-1 DR protection of this solution.
  • Remote DR center: It is usually built hundreds or thousands of kilometers away from the production center to protect against regional disasters. Data is periodically, asynchronously replicated to the remote DR center. This is the level-2 DR protection of this solution.

You can use Huawei Cloud SDRS and CBR to deploy two-site three-center DR. This solution responds quickly to small-scale failures and large-scale natural disasters, protecting service data and guaranteeing service continuity.

What Is SDRS?

Storage Disaster Recovery Service (SDRS) provides DR services for many cloud services, including Elastic Cloud Server (ECS), Elastic Volume Service (EVS), and Dedicated Distributed Storage Service (DSS). By leveraging technologies, such as storage replication, data redundancy, and cache acceleration, SDRS offers high data reliability and service continuity for users.

What Is CBR?

Cloud Backup and Recovery (CBR) enables you to back up ECSs, Hyper Elastic Cloud Servers (HECSs), Bare Metal Servers (BMSs), EVS disks, SFS Turbo file systems, a single or multiple files and databases on your local hosts, and on-premises VMware virtual environments with ease. In case of a virus attack, accidental deletion, or software or hardware fault, you can restore data to any point in the past when the data was backed up.

What Is the Two-Site Three-Center DR Solution (SDRS+CBR)?

In this solution, SDRS synchronizes data from the production site to a DR site in a different AZ of the same region for cross-AZ DR. CBR periodically backs up the production data to the other DR site in a different region for cross-region DR. The production site, cross-AZ DR site, and cross-region DR site all together form the geo-redundant architecture to meet the two-site three-center DR requirements.

The cross-AZ DR site corresponds to the intra-city DR center, and the cross-region DR site corresponds to the remote DR center.

In this document, the two-site three-center DR solution (SDRS+CBR) is called the SDRS+CBR DR solution for short.

In the SDRS+CBR DR solution, the solution is deployed in two regions and involves three sites: a production site, a cross-AZ DR site in the same region (cross-AZ DR center for short in the following document), and a cross-region DR site.

In this solution, SDRS synchronizes data from the production site to the cross-AZ DR site. CBR periodically backs up the production data to the cross-region DR site.

Application Scenarios

The SDRS+CBR DR solution is suitable if you use ECSs and EVS disks to run workloads.

Solution Advantages

Compared with two-site solutions (the cross-AZ DR solution and cross-region DR solution), the SDRS+CBR DR solution combines the advantages of both two-site solutions and can protect against both AZ-level failures and regional disasters.

  • Disk data consistency is ensured during DR switchover if an AZ-level failure occurs.
  • If a regional disaster occurs, this solution can restore the service data to the latest backup time point to ensure minimal data loss.

Table 1 compares the three types of DR solutions.

Table 1 Solution comparison

DR Solution

Handle AZ-Level Failures

Perform AZ-Level Failover to Restore System to the State Before It Broke Down

Handle Regional Disasters

Cross-AZ DR (SDRS)

×

Cross-region DR (CBR)

×

Two-site three-center DR (SDRS+CBR)

Constraints

  • Cloud-Init must be installed on Linux ECSs, and Cloudbase-Init on Windows ECSs. After you perform a failover for the first time and start the servers, the system will run Cloud-Init/Cloudbase-Init to inject the initial data. The passwords or key pairs for logging in to the production and DR servers will change.
  • If you use SDRS to set up DR, DR servers or disks can be restored only when protection is disabled for the servers.

For details about SDRS usage restrictions, see SDRS Usage Restrictions.

For details about CBR usage restrictions, see CBR Usage Restrictions.