With automatic scaling, you can configure the software to deploy access nodes in Amazon regions only when they are needed, and then decommission the access nodes soon after you use them. The software uses power management (enabled by default) to power off access nodes after you stop using them. Automatic scaling, combined with power management, can reduce the cost of using Amazon access nodes to back up VMs.
You can also perform backup copy operations using automatic scaling.
When automatic scaling is configured for an Amazon hypervisor, Commvault automatically deploys access nodes that run on Amazon Linux (64-bit ARM) or Red Hat Enterprise Linux (64-bit X86), by cloning from the Commvault Cloud Data Manager Amazon Machine Image (AMI) that corresponds to the Commvault feature release for the deployment. The Cloud Data Manager AMI is deployed for each region that has instances that are backed up for the hypervisor.
The number of access nodes that are required increases or decreases based on the amount of data you back up and the settings for automatic scaling.
The Cloud Data Manager AMI is a machine image for Amazon Linux. No Windows AMIs are available. Commvault updates the image for each feature release, including applying any required security patches. You cannot log on to the instances that are deployed from the AMI.
You can use an additional setting to configure the autoscale access node to use a custom machine image instead of the Commvault machine image. In the additional setting Value, enter the location of the machine image file on your computer, for example C:\Documents\Images\package.ova.
For the AWS autoscale setting, you can use the CVAutoScaleAMIForProxy additional setting. For more information on using additional settings, see Additional Settings.
Required CommCell Components
The following CommCell components are required for automatic scaling:
An on-premises access node or an Amazon Elastic Compute Cloud (EC2) access node
An empty server group where Amazon access nodes are created and decommissioned
A network topology, so that the access node and the server group can communicate
A hypervisor for the Amazon IAM Admin account
Access Node Scaling
The software automatically scales out the number of access nodes based on the following characteristics:
The Recovery Point Objective (RPO)
The throughput rate
The number of data readers that are configured for the VM group
The Maximum number of access nodes setting in the Auto scale settings dialog box
Important: To increase the number of access nodes, change the Maximum number of access nodes setting. The maximum number of access nodes is initially determined by the maximum number of streams. For more information, see Configuring a Hypervisor for Automatic Scaling.
Environments That Have Multiple Amazon Accounts
For environments that have multiple Amazon accounts (such as in multi-tenant or managed service provider environments), you can add a hypervisor for the Amazon IAM Admin account, and then configure additional hypervisors for each Amazon tenant account. Users can perform cross-account backup operations using automatic scaling. In this case, Amazon tenant user backup operations run on the Amazon IAM Admin account.
Access Node Expiration
Automatically scaled access nodes shut down after they are idle for 30 minutes. If an access node is shut down when a job request is initiated, then the access node is powered on and used. If no activity occurs for an additional 30 minutes on the access node that is shut down, then the access node is deleted from both the Command Center and the AWS portal.
To troubleshoot failed backup operations, you can view the log files for automatic scaling at the following location on the CommServe server: Commvault\ContentStore\iDataAgent\JobResults\CV_JobResults\AutoScale.