Copy the source machine's file system

Modernization of an application component requires creating a copy of the source machine's file system.

This page describes the steps required to copy the source machine's file system along with some specifications for reducing the size of the copied file system.

Create a local copy of the Linux source machine file system

The Migrate to Containers CLI supports access to source machines using either direct SSH connection or by using gcloud.

Direct SSH

To copy using direct SSH access, run the following command:

./m2c copy ssh [USERNAME@]HOSTNAME \
  -i PATH_TO_PRIVATE_KEY \
  -o OUTPUT_FILESYSTEM_DIR

Replace the following:

USERNAME: your username
HOSTNAME: the hostname
PATH_TO_PRIVATE_KEY: the path to private key
OUTPUT_FILESYSTEM_DIR: the path for the output directory on your local machine where you want to copy the source machine's file system

gcloud

If the source machine has a public IP, run the following command:

./m2c copy gcloud \
  -p PROJECT_ID \
  -z ZONE \
  -n VM_NAME \
  -o OUTPUT_FILESYSTEM_DIR

Replace the following:

PROJECT_ID: the project ID
ZONE: the zone of your VM
VM_NAME: the name of the source VM
OUTPUT_FILESYSTEM_DIR: the path for the directory output on your local machine where you want to copy the source machine's file system

If the source machine doesn't have a public IP, you can do one of the following:

If you are running the Migrate to Containers CLI from a machine in the same internal network, Use the --internal-ip flag.
If you are running on a different network, use the --tunnel-through-iap flag.

For more information about these flags, see the gcloud documentation.

After the copy is complete, a copy of the source machine's file system is available in the specified output directory.

Reduce the size of the copied file system

Depending on the source machine, the copied file system might be very large. The larger the copy, the longer it takes for each operation to complete and, if there's not enough space, the copy operation fails.

The Migrate to Containers CLI applies filters to reduce the size of the copy, but with the given knowledge of the application component, you can modify the filters to further reduce the size of the copy.

When you run the copy command again with the modified filters, the contents of the output directory are updated according to the changes you applied in the filter file.

To modify the default filters, follow these steps:

Get a list of the default filters:

./m2c copy default-filters > filters.txt

Edit the file filters.txt to remove irrelevant directories. The file uses the rsync filter rule format, as specified in the rsync documentation page.

Use the --filters flag to specify the filter file with the copy command.

Direct SSH

./m2c copy ssh [USERNAME@]HOSTNAME \
  -i PATH_TO_PRIVATE_KEY \
  -o OUTPUT_FILESYSTEM_DIR \
  --filters filters.txt

gcloud

./m2c copy gcloud \
  -p PROJECT_ID \
  -z ZONE \
  -n VM_NAME \
  -o OUTPUT_FILESYSTEM_DIR \
  --filters filters.txt

Create a local copy of the Windows source machine file system

For Windows source machines, you need to copy the disk images in VHD or VHDX format for the migration process. However, Migrate to Containers CLI does not support exporting disk images from your source platform to your local machine.

To export a Compute Engine virtual machine (VM) instance disk, first create an image from the disk and then export the image to the local machine.
To export an image from Compute Engine, perform the following steps:
1. Export the image to Cloud Storage using the gcloud compute images export command:
```
gcloud compute images export \
  --export-format vhdx \
  --destination-uri DESTINATION_URI \
  --image IMAGE_NAME
```
  Replace the following:
  - DESTINATION_URI: the Cloud Storage URI destination for the exported image file.
  - IMAGE_NAME: the name of the disk image to export.
2. Download the image on your local machine:
```
gcloud storage cp DESTINATION_URI LOCAL_PATH
```
  Replace the following:
  - DESTINATION_URI: the Cloud Storage URI destination of the exported image file
  - LOCAL_PATH: the path to the local folder where you want to download the image
For more information, see Export a custom image to Cloud Storage.
For more information on exporting Amazon EC2 images, see Exporting an instance as a VM using VM Import/Export.

Note: When you create the JSON file for exporting an image, omit the ContainerFormat field, and set DiskImageFormat to "VHD".
For more information on exporting Azure Cloud Compute images, see Download a Windows VHD from Azure.
To export VMware VM disk images, first complete the steps to export an image to a VMDK file. Then, convert the VMDK file to a VHD file using third-party tools such as qemu-img.

What's next

Learn how to create a migration plan.