MaxCompute storage disaster recovery extends data storage services from a single zone to three zones within the same city. By leveraging physical isolation and low-latency network connectivity, it provides real-time data synchronization and fault isolation across data centers, preventing service interruptions due to single data center storage failures and bolstering the resilience of customer businesses.
Feature introduction
To use this feature, access the application for trial use and apply for storage disaster recovery on the New Feature Trial Application page. For more information, see Apply for trial use of new features.
MaxCompute storage disaster recovery converts existing data into redundant storage across three availability zones within the same city, and synchronously writes incremental data to these three availability zones, thereby achieving disaster recovery at the data center level. When catastrophic events such as carrier network failures, power supply systems issues, and IDC infrastructure issues render a particular data center unavailable, the storage disaster recovery mechanism ensures uninterrupted data read and write services and no data loss, thus meeting the RPO=0 data recovery requirements. Additionally, this disaster recovery mechanism reduces business downtime, complies with industry regulations, and enhances user experience.
Scenarios
Finance industry
Storage disaster recovery ensures that the financial services of banks can continuously analyze and process business transaction data, and avoid service interruptions caused by data center failures.
Critical infrastructure
It ensures the continuous operation of data analysis systems for electricity, water supply, and transportation, guaranteeing that critical information services relied upon by the public do not experience interruptions due to data center failures.
Limits
Storage disaster recovery is supported only in China (Shenzhen), China East 2 Finance, and China (Hong Kong).
Metadata, user permissions, package authorization, standard tables, and Delta tables within projects are supported for storage disaster recovery. Resources are not supported.
Billing
MaxCompute is billed based on the storage disaster recovery model. For billing details, see Storage disaster recovery fee (pay-as-you-go).
Procedure
Log on to the MaxCompute console and select a region in the upper-left corner.
In the left-side navigation pane, select .
On the Intra-zone Disaster Recovery Management page, click Enable Intra-zone Disaster Recovery.
In the dialog box, select the MaxCompute project for storage disaster recovery and select the checkbox.
Click OK.
The project data starts the preparation process for storage disaster recovery, and migrate from a single zone to three zones. This process takes two days, after which the project is equipped with storage disaster recovery capabilities.
NoteJob operations remain unaffected during the storage disaster recovery preparation, ensuring business continuity.
If historical table partition data is being written in a streaming manner during the preparation, the data preparation task for the storage disaster recovery waits until the write operation is completed before starting. We recommend that you periodically switch to new partitions for data writing to ensure that all tables and partitions complete the storage disaster recovery transformation.
Related operations
Disable disaster recovery
On the Intra-zone Disaster Recovery Management page, find the target project and click Disable Disaster Recovery in the Actions column. Follow the prompts to enter the project name and click OK.
After cancellation, project data revert to local storage in a single zone.
Disabling disaster recovery is a high-risk operation that causes the project to immediately lose its disaster recovery capabilities.