Database geo-disaster recovery is to deploy a standby database system in a remote location and use the automatic failover mechanism to ensure that when the primary data center is interrupted due to disasters (such as natural disasters and hardware failures), the business can be quickly recovered and data loss (RPO) and downtime (RTO) can be minimized.
In addition to the need for local high availability, higher requirements are also put forward for remote disaster recovery capabilities
Due to the requirements of data localization, it is necessary to build disaster recovery centers across different cloud platforms and regions to avoid the risk of cross-border data transmission
Due to its high reliance on industrial control systems and real-time data processing, seamless disaster recovery was required across IDC and multi-cloud environments.
scheme | principle | merit | shortcoming |
Storage-layer replication | Cross-cloud block-level replication is achieved through the storage replication function of cloud vendors | Low performance loss, suitable for large-scale databases | Depends on the services of specific cloud vendors, has poor cross-cloud compatibility, and has limited private cloud deployment |
Offline backups | Periodically transfer the full backup file to an off-site location | 成本低，适合非核心业务 | Long recovery time (RTO>1 hour) and high risk of data loss (RPO=backup interval) |
Real-time dual write | The application layer writes data to both public and private cloud databases (e.g., dual-write SDKs) | Achieve RPO=0 without the need for additional replication tools | Application transformation costs are high, and cross-cloud network latency can lead to data inconsistencies |
Database native synchronization | Use a distributed database to deploy a single logical cluster across multiple geographies with built-in mechanisms for data replication | 支持跨地域数据冗余，简化了数据管理和一致性维护;可提供较低的 RPO/RTO 指标 | Network latency may affect performance, and cross-region leased lines are costly. The fluctuation of long-distance transmission network makes it difficult to ensure data consistency |
Incremental replication | Perform incremental data replication across clouds and regions by parsing database logs | 支持混合云准实时复制（RPO≈0）， RTO<60 秒;自动化运维降低复杂度 | It takes time to get acquainted with the tool ecosystem |
01
KubeBlocks can be quickly deployed in the form of Database-as-a-Service (DBaaS) in both private and public cloud environments, supporting a wide range of database types, including MySQL, PostgreSQL, Redis, MongoDB, MSSQL, Oracle, and Xinchuang databases such as Dameng and Jincang. Provides the database cluster lifecycle management function, simplifies the backup and restoration process, and ensures that database resources can efficiently respond to business needs.
02
The cross-cloud synchronization engine ensures real-time data replication between disaster recovery instances and production instances, significantly reducing business interruption time to the second level and enhancing business continuity. The engine uses a variety of technologies to optimize the performance of long-distance data transmission, including resumable data transmission, network compression, etc., and supports a variety of secure connection methods, flexibly adapting to various data sources in complex network environments, and ensuring the security and reliability of data transmission.
03
By providing visualized disaster recovery monitoring, switchover operations, and drill tools, the O&M difficulty is reduced and the success rate of failover is improved. Periodically check data consistency to ensure the validity and reliability of disaster recovery data, so that services can be quickly and successfully restored in the event of a failure.
MySQL
5.1/5.5/5.6/5.7/8.0
PostgreSQL
10/11/12/13/14/15/16
Redis
3.0/3.2/4.0/5.0/6.2/7.0
MongoDB
3.6/4.0/4.2/4.4/5.0/6.0
SQL Server
2008/2008R2/2012/2014
/2016/2017/2019/2022
Oracle
11g/12c/18c/19c/21c/23ai
Dameng
8.1
Golden Warehouse
v8/v9
The cloud ape data remote disaster recovery solution supports private cloud deployment, ensures the localization of enterprise core data storage, meets the security compliance requirements of "data does not leave the domain", and is suitable for industry scenarios that are highly sensitive to data sovereignty, such as finance and government affairs.
Based on incremental replication technology, the cloud ape data remote disaster recovery solution realizes real-time cross-cloud replication to ensure business continuity, with an RPO approaching 0 and an RTO controlled within 60 seconds, effectively coping with data center-level faults and improving system disaster recovery capabilities.
It supports one-way or two-way replication of multi-database engines, adapts to multiple network environments, and has the capabilities of resumable transmission, concurrent transaction processing, and DDL synchronization, significantly improving the efficiency and stability of cross-environment data migration.
It provides a unified console to centrally manage database resources in a multi-cloud environment, and integrates functions such as automatic O&M, backup and restoration, monitoring and alarming, and disaster recovery drills, significantly reducing the O&M complexity under the hybrid cloud architecture.
