How do I migrate my Amazon Redshift cluster to an RA3 node type?

Last updated: 2020-10-15

I want to migrate my Amazon Redshift cluster from an existing node to an RA3 node type. How can I upgrade my cluster configuration to the RA3 node type?

Short description

RA3 nodes allow you to optimize your data by scaling and paying for compute and managed storage. Based on your use case, you can migrate your existing Amazon Redshift cluster to the ra3.4xlarge or ra3.8xlarge node types. If you create an RA3 cluster, note that the minimum number for RA3 clusters is two nodes. For more information about RA3 node types, see Upgrading to RA3 node types.

To migrate your cluster from an existing node type to RA3, use one of the following methods:

  • Elastic resize: This is the most efficient way to change the instance type and update the nodes in your Amazon Redshift cluster. The cluster endpoint doesn't change and the downtime during resize is minimal.
  • Snapshot and restore method: Choose the snapshot and restore method if elastic resize is unavailable (from a mismatch between slice and node count). Or, use this method to minimize the amount of time that it takes to write to your production database.
    Note: Data written to the source cluster after the snapshot is taken must be copied over manually to the target cluster afterwards.
  • Classic resize: Choose the classic resize method if it's the only option available. For single node clusters, only a classic resize can be performed to convert the cluster into a multi-node cluster.

Note: When you resize your Amazon Redshift cluster, you can schedule the resize. Use the create-scheduled-action command or CreateScheduledAction API to automate cluster resizing in Amazon Redshift.

Resolution

Elastic resize

With Amazon Redshift elastic resize, you slice your cluster in half or double the node count. Amazon Redshift presents slice mapping options that are applicable to your configuration. Amazon Redshift does not consider data size during a resize.

To check the available node migration options, choose the Nodes dropdown list in the AWS Management Console. If your current configuration supports elastic resize, then use elastic resize to upgrade your cluster with the RA3 node type. The minimum node count required to perform an elastic resize is two.

Note: Some configuration changes don't allow an elastic resize. If elastic resize is unavailable as an option in your AWS Management Console, use the snapshot restore method.

Before you perform an elastic resize, determine the possible node configurations for different node types using the following:

  • The DescribeNodeConfigurationOptions API, which returns properties of possible node configurations such as node type, number of nodes, and disk usage for the specified action type.
    -or-
  • The describe-node-configuration-options AWS Command Line Interface (AWS CLI) command, which returns possible node configurations. Node configuration settings include node type, number of nodes, and disk usage for the specified action type.
    Note: The AWS CLI doesn't offer node configuration options for slice mapping.

For example:

aws redshift describe-node-configuration-options --action-type resize-cluster --cluster-identifier your-redshift-clusteridentifier

To use elastic resize to change your cluster configuration in Amazon Redshift, perform the following steps:

1.    Sign in to your AWS Management Console.

2.    Open the Amazon Redshift console.

3.    Choose Clusters.

4.    Select your Amazon Redshift cluster.

5.    Choose Actions.

6.    Choose Resize.

7.    Select Elastic Resize as your resize type.

8.    Select the node that you want to migrate to for New cluster configuration.

9.    Select the number of nodes based on node type.

10.    Choose Resize cluster now to trigger the resize. When the resize completes, the status of your cluster changes from Available to Modifying.

Snapshot and restore

If elastic resize is unavailable to you in your Amazon Redshift console, use the snapshot and restore method. You can use the DescribeNodeConfigurationOptions API or describe-node-configuration-options AWS CLI command to obtain the possible node configurations for different node types. Then, restore your cluster to the original configuration.

For example:

aws redshift describe-node-configuration-options --action-type restore-cluster --snapshot-identifier your-snapshot-identifier

To use the snapshot and restore method to change your cluster configuration, perform the following steps:

1.    Sign in to your AWS Management Console.

2.    Open the Amazon Redshift console.

3.    Choose Clusters.

4.    Choose Snapshots.

5.    Select the most recent snapshot of the source cluster.

6.    Choose Restore from Snapshot.

7.    Select the node type.

8.    Select the number of nodes.

9.    (Optional) Under Additional Configuration, modify the database configurations and default configurations (such as your VPC, parameter group, monitoring, and backup).

10.    Choose Restore cluster from snapshot.

To retain the same endpoint as your source cluster, perform the following steps:

Important: Before you begin, you must create an Amazon Redshift cluster.

1.    Delete the old Amazon Redshift cluster.

2.    Select the new cluster.

3.    Choose the Actions dropdown menu.

4.    Choose Modify.

5.    Specify the same Cluster Identifier as your deleted cluster.

6.    Choose Modify Cluster.

You can also use the ModifyCluster API or modify-cluster AWS CLI command to rename your Amazon Redshift cluster. For more information, see Renaming clusters.

Classic resize

If elastic resize is unavailable to you in your Amazon Redshift console, use the classic resize method to update your node configurations. A classic resize depends on the amount of data and node count in your cluster. The resize can take anywhere from several hours to several days. It's a best practice to perform a classic resize if it is your only option for upgrading your node configuration.

Note: In some cases (such as single node clusters), the classic resize is your only option to update your node configuration.

To use a classic resize to change your node configuration, perform the following steps:

1.    Sign in to your AWS Management Console.

2.    Open the Amazon Redshift console.

3.    Select the cluster that you want to modify.

4.    Choose the Actions dropdown menu.

5.    Choose Resize.

6.    Choose Classic Resize.

7.    Under New cluster configuration, select your preferred node type and nodes for migration.

8.    Choose Resize cluster now to begin the classic resize. Your Amazon Redshift cluster remains in the read-only mode until the resize operation completes.


Did this article help?


Do you need billing or technical support?