Cross-Cloud DR

If critical application systems are deployed in an IDC or another cloud environment and need to be backed up to Huawei Cloud for disaster recovery (DR), a cross-cloud DR solution is essential to ensure high availability. Assuming the systems in the IDC or other cloud already have 99.9% availability, using Huawei Cloud DR can bring it up to 99.99%.

A typical architecture for cross-cloud applications includes a frontend stateless application layer and a backend database layer. The frontend application layer uses VMs or containers (using services like Huawei Cloud CCE), while the backend database layer uses MySQL databases (like you get with Huawei Cloud RDS for MySQL) to support cross-cloud DR.

The recommended deployment on Huawei Cloud is as follows.

Item	Solution
Redundancy	Deploy cloud services like ELB, CCE, DCS, Kafka, RDS, and DDS, to ensure high availability.
Backup	Enable automated backup for RDS and DDS databases. When a data fault occurs, the latest backup can be used to restore data, meeting availability requirements.
DR	Deploy applications across AZs. If an AZ fails, applications can be automatically recovered. Use cross-cloud DR to switch workloads to Huawei Cloud quickly if an IDC or another cloud environment fails.
Monitoring	Monitor and check site statuses and instance workloads. If a site fails or a CCE, DCS, Kafka, RDS, or DDS instance is overloaded, an alarm is reported.
Auto scaling	Use CCE clusters with auto scaling of workloads.
Change error prevention	Use canary or blue-green deployment for software updates. Canary and blue-green deployments can be completed automatically and rolled back if an update does not go smoothly.
Emergency recovery	Develop an emergency response system and designate emergency personnel to make decisions quickly and recover services. Provide solutions for common application and database issues, as well as upgrade and deployment failures. Conduct periodic drills to promptly identify and address potential problems.

A typical deployment architecture is as follows.

The architecture has the following features:

The application system uses a multi-layered architecture featuring stateless applications and stateful databases or VMs.
The application system is deployed either in an IDC or on another cloud platform, as well as on Huawei Cloud. On Huawei Cloud, the system is distributed across AZs to facilitate active-active DR at the application level between different data centers. Data from other clouds can be synchronized to Huawei Cloud in real-time. If there is a failure in the IDC or another cloud, the primary/standby DR solution enables workloads to be rapidly switched over to Huawei Cloud.
- Access layer (external DNS servers and API gateways): This layer utilizes external DNS servers for domain name resolution. If the IDC or another cloud environment fails, service traffic is manually redirected to Huawei Cloud.
- Application layer (load balancers, application software, and VMs or physical hosts): Load balancers are employed for stateless applications to detect faults and distribute loads. On Huawei Cloud, Auto Scaling (AS) monitors loads and triggers automatic scaling. For stateful applications, Storage Disaster Recovery Service (SDRS) facilitates cross-cloud VM data replication and DR switchover. In addition, Cloud Backup and Recovery (CBR) is used for automated data backups.
- Middleware layer: Redis and Kafka clusters are deployed across AZs for high availability.
- Data layer: MySQL databases are designed for high availability. Data Replication Service (DRS) is used to enable cross-cloud database replication and DR switchover. Data is automatically backed up on a regular basis. If any data is lost, workloads can be quickly restored.
To ensure data reliability, database data is automatically backed up periodically, enabling quick restoration in the event of data loss.