Feature Introduction
When the data storage space of nodes in a cluster becomes insufficient due to business growth, a scale-out is required. This document describes how to scale out cloud data disks in the EMR console.
Note:
Only cloud data disks support scale-out operations, while local disks do not.
Pod and shared resource nodes do not support scaling out cloud disks.
It is recommended to create a snapshot of a cloud disk before scaling out the disk to prevent file system damage caused by maloperations.
To prevent data loss, disks can only be scaled out, but not scaled in.
When scaling out cloud data disks for multiple nodes in batches, you can only operate nodes with the same billing mode, availability zone (AZ), resource type, and node type.
Operation Steps
1. Log in to the EMR console, then click the corresponding Cluster ID/Name in the cluster list to go to the cluster details page. 2. Go to the resource management page, and select the single-node operation or batch operation solution as needed:
2.1 Single-node scale-out of cloud data disks: Select the node that requires a scale-out. Choose More > Disk Adjustment > Scale out Cloud Data Disks in the operation column of the list to go to the cloud data disk scale-out page.
2.2 Multi-node batch scale-out of cloud data disks: Choose Disk Adjustment > Scale out Cloud Data Disks in the More Operations header to go to the cloud data disk scale-out page.
3. In the Select Target Hard Disk pop-up window, select the cloud disk that requires a scale-out, then click Next.
4. In the Adjust Capacity step, specify the Unified Target Capacity (the value should be greater than the current capacity). After the adjustment, the capacity of the selected data disk will be resized to the specified capacity.
5. After confirming that the information is correct, click Confirm to start the cloud data disk scale-out.
6. After the disk scale-out is completed, the partition and file system will be scaled out automatically, eliminating the need for manual updates.