Cloud Server High Availability Video
Cloud Server High Availability (CSHA) provides HA protection between data centers in the same city for ECSs. If the production center fails in a disaster, the protected ECSs can be automatically or manually switched to the DR center.
CSHA uses the storage-layer HyperMetro technology and supports zero RPO. The computing performance of ECSs is not affected during the replication.
Data consistency allows tenants to perform consistent DR protection for all volumes of one ECS or a group of ECSs. The DR startup sequence of ECSs and CSHA instances can be defined.
CSHA provides HA protection for ECSs. If the production center is faulty, services on the protected ECS can be automatically or manually switched to the DR center.
Easy to Use
For visualized service application and DR configuration, there is no need for application reconstruction. Faults are automatically recovered with only one click.
Two DCs are respectively deployed in two AZs in a region. One is the production center, and the other is the DR center. The third DC functions as the quorum site and provides the arbitration service for switchover of the cloud platform and storage. Service systems protected by CSHA instances are deployed on ECSs in the production center.
When the DC in the production center is faulty or planned to be shut down, you can manually switch the ECSs protected by CSDR to the DR center through one click. RPO ≈ 0; RTO: dozens of minutes.
This solution is suitable for scenarios where the web, applications, and databases are deployed as a whole on ECSs. The service system does not support disaster recovery, which will be provided by a cloud platform for ECSs.
Tenants can apply for CSHA protection for ECSs without application reconstruction.
When the production center is faulty, active-active storage replication ensures zero data loss.
The failure of a single storage device does not lead to service interruption or data loss (RPO=0). During the replication of storage data, the computing performance of ECSs will not be affected.
Key management nodes (ManageOne, BCManager eReplication, and FusionSphere OpenStack) that support fault recovery are remotely deployed at different sites, and are interconnected with third-place arbitration nodes to implement automatic switchover in scenarios such as site faults and link faults. Other management nodes support cross-site DR and manual switchover when a site is faulty.
Tenants can perform consistent active-active storage protection for all disks of one ECS or a group of ECSs.
The cross-site active-active deployment at the storage layer is provided based on IaaS, without being aware of applications in the ECS.
The automatic and manual switchovers are both supported
In automatic switchover mode, ECSs will be automatically switched over between sites when one site fails. When the site recovers, reprotection will automatically start.
In manual switchover mode, the administrator can perform one-click fault recovery for affected ECS protected instances when some sites are faulty.