VM Conversion Using the Import Method

Updated

You can convert a VMware, Azure, or Hyper-V VM to an Amazon EC2 instance using the AWS VM Import/Export method.

Requirements

Windows Source VMs

  • The pagefile must be located on the C: drive.

Linux Source VMs

  • fstab entries must use UUID instead of the device name.

  • You cannot use a Linux Access node for the AWS VM Import/Export method.

Disk Space Requirements

  • Ensure that there is a minimum of 2 GB free space on the OS disk for installing drivers and other software.

Network Requirements

  • The access node VM must be able to connect to ec2.amazonaws.com. For more information, see Configuring a Firewall for an Access Node in the Cloud.

Configuring Source VMs for Import

  • When converting a VM to a nitro instance type, nitro drivers must be pre-installed on the guest. For more information, see Preparing the Source VM.

  • When replicating a VM to a nitro instance type, nitro drivers must be pre-installed on the guest. For more information, see Preparing the Source VM.

Import Process

Supported Operating Systems

For information about Amazon's supported VM operating systems, see VM AWS Import/Export Requirements.

Operating systems can be imported to and exported from Amazon EC2. For more information about whether a Region is enabled by default, see "Available Regions" in the Amazon EC2 User Guide for Linux Instances.

Supported Volume Types and File Systems

For information about Amazon's supported volume types and file systems, see "Volume types and file systems" in Amazon EC2 User Guide.

Process

  1. Upload the virtual VM disks to S3 as pages.

    The naming convention for the temporary S3 bucket that is used to hold VHD files is gx-restore-region_name-account_id. You can override the default naming convention for S3 buckets by adding the AmazonBucketForImport additional setting to the VSA access node.

    For instructions about adding the additional setting from the CommCell Console, see Add or Modify an Additional Setting.

    Note that the S3 pages are retained as integrity snapshots for the next incremental replication.

  2. Merge the S3 pages uploaded for each disk to form a single VHD.

  3. Import the OS disk as an AMI and additional disks as snapshots.

  4. Launch an Amazon instance from the imported AMI and attach the additional disks (volumes) from the imported snapshots.

  5. Delete the AMI and snapshots from AWS and the merged VHDs from the S3 bucket.

Note: If you are using the Import method when you perform a full instance restore operation for AWS, convert a VM to an AWS instance, or replicate a VM to AWS by using the Import method. Commvault creates an S3 bucket for the destination automatically. The bucket is named gx-restore-region_name-account_id, where region_name is the name of the AWS region and account_id is the AWS account for the virtualization client.