DocsICP FilingConsole
官方文档
首页

Start a restore job

更新时间：
复制 MD 格式
产品详情
我的收藏

You can use the restore feature to restore data at the tenant, database, and table levels.

Background information

You can restore data from existing backup sets. These sets include physical backups and logical backups.

Note

To enable logical backup and restoration, contact OceanBase technical support.

When you start a restore job, you restore data from a source tenant to a target tenant. For example, if you restore data from tenant A to tenant B, tenant A is the source tenant and tenant B is the target tenant.

OceanBase Database supports local backup and restoration. If you enable geo-redundancy, you can also restore data from geo-redundant backup files.

Backup and restoration has some limitations. Common limitations include the following:

  • The target cluster must be in the `Running` state or have its whitelist being modified. Otherwise, you cannot restore data to it.

  • In a primary/standby cluster setup, a standby cluster cannot be a restore target. For the primary cluster, only V4.x versions can be a restore target. Versions earlier than V4.0.0 are not supported.

  • The source and target instances have mismatched types. You cannot restore data between cluster instances and Serverless instances.

  • The source and target instances have mismatched product series. You can only restore a relational database to another relational database. The same applies to non-relational databases.

  • You cannot restore a backup from a later version of OceanBase Database to an earlier version.

  • Generally, you can restore a backup from an earlier version of OceanBase Database to the same or a later version. However, some special scenarios are not supported.

    • You cannot restore data backups from OceanBase V2.x or V3.x to a V4.x cluster instance.

    • You cannot restore data backups from OceanBase V4.0 to a V4.1 cluster instance.

Cluster instance restoration

  1. Log on to the OceanBase console .

  2. In the navigation pane on the left, click Instance List and select the cluster instance that you want to restore.

  3. On the instance details page, click Backup and Restoration in the navigation pane on the left.

  4. Click Start Restore Job. On the Start Restore Job page, select the object to restore and configure the parameters. You can restore data at the tenant, database, or table level. The parameters are described in the following sections.

  5. Click OK to start the restore job. In the dialog box that appears, confirm the resource status of the restore object.

    1. If the restore object has sufficient resources, the dialog box appears as shown in the following figure.

    2. If the restore object has insufficient resources, the dialog box displays a message indicating that the target instance has low storage space. The dialog box appears as shown in the following figure.

  6. In the dialog box, click Confirm. You can then view the status on the Restore Jobs page of the target cluster. For more information about restore job details, see Data Restoration.

Tenant restore job

Parameter

Description

Restore Source

Select the source of the data to restore.

Note

If you select Geo-redundancy, the current cluster is the target cluster for restoration. If another cluster has a geo-redundancy region that is the same as the region of the current cluster, you can restore data from the geo-redundant backup files. For example, cluster A is in Hangzhou and its geo-redundancy region is Shanghai. Cluster B is in Shanghai. You can use the geo-redundant backup files of cluster A in Shanghai to restore a tenant of cluster A to cluster B.

Restore Method

Select whether to restore by point in time or by backup set.

  • If you restore by backup set, you must select a backup set and a source tenant.

  • If you restore by point in time, you must select a source tenant and a restore point.

Source Cluster

Select a cluster that has geo-redundant backup files in the current region. This parameter is required when the restore source is Geo-redundancy.

Source Tenant

Select the tenant whose data you want to restore.

Restore Point

Select the point in time to which you want to restore the data.

Backup Set

Select the backup set to restore.

Target Instance

Select the instance to which you want to restore the data. If the restore source is Geo-redundancy, the current cluster is the target instance.

Note

A grayed-out instance cannot be restored to. Hover your mouse pointer over the cluster name to view the reason.

Restore Whitelist

Select whether to restore the whitelist of the source tenant.

Note

When you restore to another cluster, you must select a whitelist configuration.

Target Tenant Name

Select the tenant to which you want to restore the data.

Number of Resource Units

Set the number of resource units for the tenant. Each unit contains three nodes. Each unit you add increases the node resources by three.

Allocated Resources

Set the CPU and memory size for the tenant on a single node.

Note

  • Cluster instances of OceanBase Database V3.x and V4.x support the creation of 1C4GB tenants.

  • The total CPU and memory size of all tenants cannot exceed the specifications of the cluster.

  • Total available resources for a tenant = Allocated resources per node × Number of resource distribution nodes × Number of replicas.

  • For OceanBase Database V4.x, you can further adjust the log disk size in Advanced Settings.

VPC

Select the virtual private cloud (VPC) ID of the network where the ECS instance is located.

If no suitable VPC is available, create one as prompted on the page. For more information, see What is a VPC?.

Primary Zone

Select the primary zone for the target tenant.

vSwitch

Select the vSwitch ID of the network where the ECS instance is located. The number of available IP addresses for the vSwitch is displayed.

If no suitable vSwitch is available, create one as prompted on the page. For more information, see Work with vSwitches.

Notes (Optional)

The notes cannot exceed 30 characters in length.

Backup and restore

Database restore job

Database restoration supports logical database restoration. The parameters are described in the following table.

Note

To enable logical backup and restoration, contact OceanBase technical support.

Parameter

Description

Restore Method

Database restore jobs only support restoration from a backup set.

Backup Set

Select the backup set to restore.

Target Instance

Select the instance to which you want to restore the data. If the restore source is Geo-redundancy, the current cluster is the target instance.

Note

A grayed-out instance cannot be restored to. Hover your mouse pointer over the cluster name to view the reason.

Target Tenant

Select the tenant to which you want to restore the data.

Notes (Optional)

The notes cannot exceed 30 characters in length.

Table restore job

Table restoration supports logical table restoration and physical table restoration.

  1. Logical table restoration

    You can restore a logical table to an existing tenant. The following figure shows an example.

    Note

    To enable logical backup and restoration, contact OceanBase technical support.

    Parameter

    Description

    Restore Method

    Select to restore by backup set to perform a logical table restoration.

    Select Backup Set

    Select the backup set for table restoration.

    Table Object Selection

    Select the database and table to restore. You can also search by entering the tenant, database, or table name in the search box.

    Note

    You can select a maximum of 200 tables for a single restore job.

    Target Instance

    Select the instance to which you want to restore the data.

    Target Tenant

    Select the tenant to which you want to restore the data.

    Target Database

    Select the database to which you want to restore the data.

    Start restore job for a table

  2. Physical table restoration. You can restore a physical table to a new tenant in the cluster.

    Note

    Currently, only OceanBase Database V3.x supports physical table-level restoration.

    Parameter

    Description

    Restore Method

    Select whether to restore by backup set or by point in time.

    • If you restore by backup set, you must select a backup set.

    • If you restore by point in time, you must select a source tenant and a restore point.

    Select Backup Set

    Select the backup set for table restoration.

    Source Tenant

    Select the tenant whose data you want to restore.

    Restore Point

    Select the point in time to which you want to restore the data.

    Table Object Selection

    Select the database and table to restore. You can also search by entering the tenant, database, or table name in the search box.

    Note

    You can select a maximum of 200 tables for a single restore job.

    Target Instance

    Select the instance to which you want to restore the data.

    Restore Whitelist

    Select whether to restore the whitelist of the source tenant.

    Note

    When you restore to another cluster, you must select a whitelist configuration.

    Target Tenant Name

    Select the tenant to which you want to restore the data.

    Number of Resource Units

    Set the number of resource units for the tenant. Each unit contains three nodes. Each unit you add increases the node resources by three.

    Allocated Resources

    Set the CPU and memory size for the tenant on a single node, and the log disk size for the tenant.

    Note

    • Only tenants of OceanBase Database V4.x support setting the log disk size.

    • Cluster instances of OceanBase Database V3.x and V4.x support the creation of 1C4GB tenants.

    • The total CPU and memory size of all tenants cannot exceed the specifications of the cluster.

    • Total available resources for a tenant = Allocated resources per node × Number of resource distribution nodes × Number of replicas.

    VPC

    Select the VPC ID of the network where the ECS instance is located.

    If no suitable VPC is available, create one as prompted on the page. For more information, see What is a VPC?.

    Primary Zone

    Select the primary zone for the target tenant.

    vSwitch

    Select the vSwitch ID of the network where the ECS instance is located. The number of available IP addresses for the vSwitch is displayed.

    If no suitable vSwitch is available, create one as prompted on the page. For more information, see Work with vSwitches.

    Notes (Optional)

    The notes cannot exceed 30 characters in length.

Serverless instance restoration

  1. In the navigation pane on the left, click Instance List and select the Serverless instance that you want to restore.

  2. On the Serverless Workbench, click Backup and Recovery in the left navigation bar.

  3. Click Initiate Recovery to open the Initiate Recovery page, where you can select a recovery method and configure the related parameters.

    Parameter

    Description

    Restore Method

    Select whether to restore by backup set or by point in time.

    • If you restore by backup set, you must select a backup set and a source instance.

    • If you restore by point in time, you must select a source tenant and a restore point.

    Backup Set

    Select the backup set to restore.

    Restore Point

    Select the point in time to which you want to restore the data.

    Serverless Instance Name

    Select the target instance name.

    Note

    A Serverless instance restoration creates a new Serverless instance. This new instance is billed based on its actual computing power usage.

    Restore Whitelist

    Select whether to restore the whitelist of the source tenant.

    Note

    When you restore to another instance, you must select a whitelist configuration.

    Zone

    The destination zone for recovery.

    Note

    Currently, you can only restore to the zone of the source instance.

    Minimum Computing Power

    The minimum value in the range where the computing power of the Serverless instance automatically scales with usage.

    Maximum Computing Power

    The maximum value in the range where the computing power of the Serverless instance automatically scales with usage.

    VPC network

    Select the VPC ID of the network where the ECS instance is located.

    If no suitable VPC is available, create one as prompted on the page. For more information, see What is a VPC?.

    Primary Zone

    Select the primary zone for the target instance.

    vSwitch

    Select the vSwitch ID of the network where the ECS instance is located. The number of available IP addresses for the vSwitch is displayed.

    If no suitable vSwitch is available, create one as prompted on the page. For more information, see Work with vSwitches.

    Notes (Optional)

    The notes cannot exceed 30 characters in length.

  4. Click OK to execute the restore job. You can view the job status on the Restore Jobs page of the source instance. After the target instance is restored, you can also view the restore job on its Restore Jobs page. For more information about restore job details, see Data Restoration.

Previous：Backup policyNext：Data restore