Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions,
see Getting Started with Amazon Web Services in China
(PDF).
Amazon Aurora DSQL restore
Overview
To restore a Amazon Aurora DSQL single-Region cluster, use the Amazon Backup console or CLI to
select the recovery point (backup) you wish to restore. To restore a Aurora DSQL
multi-Region cluster, you can now use either the Amazon Backup console or CLI.
For single-Region restore, include the name, cluster encryption, and deletion
protection, then initiate the restore to a newly created cluster.
For multi-Region restore, you'll need to specify additional parameters including a
witness Region, peer Region(s), and regional configuration settings. Multi-Region restore
creates a cluster that spans multiple Amazon Web Services Regions, providing enhanced availability and
disaster recovery capabilities.
Restore Aurora DSQL single Region
cluster
You can restore an Aurora DSQL cluster to a single Region by using the Amazon Backup console or
Amazon CLI.
- Console
-
Open the Amazon Backup console at https://console.amazonaws.cn/backup.
-
Select the "Restore" button next to the recovery point you wish to
restore.
-
Configure the settings for the new cluster to which your recovery point will
be restored.
-
By default, the AMK (Amazon managed key) will be used to encrypt the
restored data. You may alternatively specify a different key.
-
Deletion protection for your Aurora clusters is enabled by
default, but unselect the box to turn off the option.
-
Review the settings; when they are satisfactory, select the
Restore backup button.
Amazon Backup will create a new Aurora DSQL cluster.
- Amazon CLI
-
Single Region restore
-
Use the CLI command aws backup start-restore-job
to restore an
Aurora cluster from the specified recovery point.
-
Include the necessary metadata for the restore job. Example:
{
"recoveryPointArn": "arn",
"tags": Map<String, String>,
"metadata": {
"regionalConfig": [
{
"region": "us-east-1
", // optional
"isDeletionProtectionEnabled": true, // optional
"kmsKeyId": "my_key
" // optional
},
]
}
}
Restore an Aurora DSQL multi-Region
cluster
Aurora DSQL multi-Region cluster restore occurs within a closed Region triplet, which
is a group of three Amazon Web Services Regions peers. Multi-Region restore requires that the Regions
you specify in the operation are contained in one triplet. For more information about
multi-Region clusters, see Configuring
multi-Region clusters.
The three triplets are:
-
US East (N. Virginia); US East (Ohio); US West (N. California)
-
Europe (Ireland); Europe (London); Europe (Paris)
-
Asia Pacific (Tokyo); Asia Pacific (Seoul); Asia Pacific (Osaka)
To complete multi-Region restore, ensure you have the following permissions:
-
backup:StartRestoreJob
-
dsql:UpdateCluster
-
dsql:AddPeerCluster
-
dsql:RemovePeerCluster
You can restore a backup of an Aurora DSQL cluster to multiple Regions using either the
Amazon Backup console or CLI commands.
If you have a backup plan with a rule that automatically creates a cross-Region copy
to one of the indicated Regions, the created copy can be used for this multi-Region
restore.
Multi-Region restore starts with your current Region. You will also need a:
-
Peer Region with an identical cross-Region copy of the recovery point in your current
Region
-
Witness Region, a designated Amazon Web Services Region that participates in multi-Region
cluster configurations by supporting transaction log-only writes without consuming storage
for the actual data. For more information about witness Regions, see Creating a multi-Region cluster.
The individual steps are shown below:
- Console
-
The Amazon Backup console now supports multi-Region restore for Aurora DSQL clusters,
providing a streamlined process for creating clusters that span multiple Regions.
For more information about multi-Region clusters, see Configuring multi-Region clusters.
-
Sign in to the Amazon Management Console and open the Amazon Backup console at https://console.aws.amazon.com/backup.
-
In the navigation pane, choose Backup vaults.
-
Choose the backup vault that contains the Aurora DSQL recovery point you want
to restore.
-
Select the recovery point you want to restore, then choose
Restore.
-
On the restore page, under Restore options, select
Add peer Regions to enable multi-Region restore.
-
Select a Peer cluster Region from the dropdown menu.
This Region must be within the same triplet as your current Region and
also must contain a cross-Region copy from the recovery point in the current
(first) Region.
-
Select a Witness Region from the dropdown menu. This
Region must also be within the same triplet.
-
Configure the Cluster settings for both the primary and
peer Region clusters:
-
For the primary cluster, configure:
-
For the peer Region cluster, configure:
-
Review your settings and choose Restore backup.
-
The console will initiate the multi-Region restore process, which creates
clusters in both Regions and automatically links them together.
- Amazon CLI
-
Multi-Region restore can now be achieved using the new orchestrated restore
metadata with Amazon Backup CLI commands. This approach simplifies the process by handling
the cluster linking automatically. For more information about creating multi-Region
clusters programmatically, see CreateMultiRegionClusters in the Aurora DSQL API Reference.
Both the primary cluster and peer cluster must be in Regions within the same
triplet. The operation will fail if the clusters are in Regions outside the
triplet.
-
US East (N. Virginia); US East (Ohio); US West (N. California)
-
Europe (Ireland); Europe (London); Europe (Paris)
-
Asia Pacific (Tokyo); Asia Pacific (Seoul);
Asia Pacific (Osaka)
Multi-Region restore through Amazon CLI using orchestrated restore
metadata
-
Create a restore job using the CLI command aws backup
start-restore-job
with the new multi-Region orchestration
metadata:
aws backup start-restore-job \
--recovery-point-arn "arn:aws:backup:us-east-1:123456789012:recovery-point:abcd1234" \
--metadata '{
"witnessRegion": "us-west-1",
"useMultiRegionOrchestration": true,
"peerRegion": ["us-east-2"],
"regionalConfig": "[{\"region\":\"us-east-1\",\"isDeletionProtectionEnabled\":true,\"kmsKeyId\":\"arn:aws:kms:us-east-1:123456789012:key/ba4b3773-4bb8-4a7a-994c-46ede70202f5\"},{\"region\":\"us-west-2\",\"isDeletionProtectionEnabled\":true,\"kmsKeyId\":\"arn:aws:kms:us-west-2:123456789012:key/ba4b3773-4bb8-4a7a-994c-46ede70202f5\"}]"
}'
The metadata structure includes:
-
witnessRegion
: The Region that will serve as the witness
for the multi-Region cluster. For more information, see Resilience in Amazon Aurora DSQL.
-
useMultiRegionOrchestration
: Set to true
to
enable multi-Region orchestration.
-
peerRegion
: An array containing the Region(s) with peer
clusters in the multi-Region cluster. For more information, see MultiRegionProperties in the Aurora DSQL API Reference.
-
regionalConfig
: An array containing configuration for each
Region:
-
region
: The Amazon Web Services Region identifier.
-
isDeletionProtectionEnabled
: Boolean flag to
enable/disable deletion protection. For more information, see CreateCluster in the Aurora DSQL API Reference.
-
kmsKeyId
: The KMS key ARN for encryption
(optional).
If regionalConfig
properties are not specified, then
default values will be applied: default encryption and
isDeletionProtectionEnabled
= TRUE
.
-
Monitor the restore job status using the aws backup
describe-restore-job
command:
aws backup describe-restore-job --restore-job-id job-12345678
-
Once the restore job completes, you can verify the multi-Region cluster
configuration using the Aurora DSQL CLI:
aws dsql describe-cluster --cluster-identifier your-cluster-id
For more information about multi-Region cluster operations, see UpdateCluster in the Aurora DSQL API Reference.
Troubleshoot Aurora DSQL restore
issues
Error: Insufficient permissions
Possible cause: If you try to copy an Aurora DSQL recovery point
into an account (cross-account copy) that has never interacted with DSQL API, you may get
a permission issue error since the DSQL service-linked role isn't set up in the
destination account.
Remedy: Attach the DSQL managed
policy that includes the DSQL service-linked role, AuroraDsqlServiceLinkedRolePolicy, to a role in the destination
account.
If you encounter any other issues with the backup or restore process, you can check the
status of your backup and restore jobs in the Amazon Backup console or using the Amazon CLI.
Additionally, you can review the Amazon CloudTrail logs for any relevant error messages or events
related to your Amazon Backup operations.
Aurora DSQL restore frequently asked
questions
-
"Can I use Amazon Backup for Aurora DSQL from the Aurora DSQL
console?"
No, you can only perform backups and restores, as well as managing backups, from
Amazon Backup console, SDK, or CLI.
-
"What backup granularity is available for Aurora DSQL? Can I backup
specific tables or databases in my cluster"
You can only back up and restore a whole Aurora DSQL cluster.
-
"Are backups of Aurora DSQL full backups or incremental
backups?"
Recovery points of Aurora DSQL clusters (backups) are full backups of your
clusters.
-
"Can I create backups for my Aurora DSQL multi-Region
clusters?"
Yes, you can create backups for each cluster in multi-Region clusters in the using
the same steps as when you create a backup of a single cluster in a single
Region.
Amazon Backup recommends as a best practice to create a cross-Region copy of your backup
in the other Region from which you plan to restore the Multi-Region cluster, as
multi-Region restore requires an identical copy of the same recovery point
[identical in this operation means the recovery points have the
same resource name and creation time].
-
"Will my restored cluster overwrite my existing
cluster?"
No. When you restore your Aurora DSQL data, Amazon Backup creates a new cluster from your
snapshots; the restored cluster won’t overwrite the source cluster.