This page describes how to configure network connectivity to SQL Server
sources on Microsoft Azure for heterogeneous SQL Server to
AlloyDB for PostgreSQL migrations with Database Migration Service. Database Migration Service
supports Microsoft Azure SQL Managed Instance and Microsoft Azure SQL Database
as source databases.

There are several methods you can use to configure the necessary
network connectivity for migrations from SQL Server sources on Azure:

- [Public IP allowlist](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#ip-allowlist#)
- [Forward-SSH tunnel](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#forward-ssh)
- Private IP connectivity. These methods require Cloud VPN or Cloud Interconnect:
  - Recommended: [Private IP connectivity with Private Service Connect interfaces](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#private-psc-interfaces)
  - [Private IP connectivity with Virtual Private Cloud peering](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#private-vpc-peering)

To learn more about source database network connectivity, see
[Source networking methods overview](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/networking-methods-source.md.txt).

## Configure IP allowlist connectivity

To use the public IP allowlist connectivity method, follow these steps:

1. Configure your source Microsoft Azure SQL Managed Instance to allow
   public network access from
   [Database Migration Service public IP addresses](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/ip-allowlists-and-regions.md.txt) for the region where you
   create the migration job.

   For more information, see
   [Change public network access](https://learn.microsoft.com/en-us/azure/azure-sql/database/connectivity-settings#change-public-network-access) in the Microsoft documentation.
2. At a later stage, when you [create the source connection profile](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/create-source-connection-profile.md.txt), do the following:
   1. In the **Define connection details** section,
      use your instance's fully qualified server name for the source database IP.

      For more information about how to view the connection details of
      your Microsoft Azure SQL Managed Instance instance, see
      [Get server connection information](https://learn.microsoft.com/en-us/azure/azure-sql/database/connect-query-content-reference-guide) in the Microsoft
      documentation.
   2. In the **Define connectivity method** section, select **IP allowlist**.

## Configure connectivity over a forward-SSH tunnel

To connect to your source database with a Secure Shell (SSH) tunnel,
follow these steps:

1. Create a virtual machine in Microsoft Azure to serve as a dedicated
   forward-SSH tunnel server. Make sure you configure it in the same
   Azure Virtual Network where your source SQL Server has
   a private IP assigned.

   For more information, see
   [Quickstart: Create a Linux virtual machine in the Azure portal](https://learn.microsoft.com/en-us/azure/virtual-machines/linux/quick-create-portal)
   in the Microsoft documentation. You don't need to create the
   web server mentioned in the quickstart.
2. [Connect to your EC2 instance](https://learn.microsoft.com/en-us/azure/virtual-machines/linux/quick-create-portal#connect-to-virtual-machine) and set up the SSH tunnel. Follow these steps:
   1. Create a separate, dedicated user account for Database Migration Service to connect as:

      ```bash
      adduser TUNNEL_ACCOUNT_USERNAME
      ```
   2. Restrict shell access for the Database Migration Service account to enhance security:

      ```bash
      usermod -s /usr/sbin/nologin TUNNEL_ACCOUNT_USERNAME
      ```
   3. Decide what authentication method you want Database Migration Service to use
      when connecting to the tunnel.

      You can use a simple password, or generate SSH keys in the
      `PEM` format that you can later upload to Database Migration Service
      when you create the source connection profile.
      - If you want to use a password, you don't need to configure anything additional. Remember what password you created for the <var translate="no">TUNNEL_ACCOUNT_USERNAME</var> account.
      - If you want to use key-based authentication, you need to generate a private-public key pair. For example, you can use the `ssh-keygen` utility:
        1. Generate the key pair:

           ```bash
               ssh-keygen -m PEM -f YOUR_KEY_NAME
               
           ```
        2. Copy the public key (`YOUR_KEY_NAME.pub`) to the `~/.ssh/` directory on your tunnel server.
        3. Save the private key. You need to upload it later to Database Migration Service when you create the source connection profile.
   4. Edit the `/etc/ssh/sshd_config` file to configure the forward-SSH tunnel to match your organization's requirements. We recommend using the following settings:

      ```bash
      # Only allow the Database Migration Service user to connect.
      AllowUsers TUNNEL_ACCOUNT_USERNAME

      # Send keep-alive packets every 60 seconds to ensure that
      # the tunnel doesn't close during the migration
      ServerAliveInterval=60

      # Optional: Force key-based authentication
      PasswordAuthentication no

      # Enables Database Migration Service to connect from a different host
      PermitTunnel yes
      GatewayPorts yes
      ```
   5. Run the `ssh` command to start the tunnel.

      Before using any of the command data below, make the following replacements:
      - <var class="edit" scope="TUNNEL_SERVER_SSH_PORT" translate="no">TUNNEL_SERVER_SSH_PORT</var> with the port number where your server is listening for SSH connections.
      - <var class="edit" scope="SOURCE_DATABASE_PRIVATE_IP" translate="no">SOURCE_DATABASE_PRIVATE_IP</var> with the private IP address of your source database. The SSH server needs to be able to reach that IP.
      - <var class="edit" scope="SOURCE_DATABASE_PORT" translate="no">SOURCE_DATABASE_PORT</var> with the port number where your source database is listening for connections. The default port number for TCP connections on SQL Server is `1433`.
      - <var class="edit" scope="USERNAME" translate="no">USERNAME</var> with the name of the user account that will be running the tunnel. This is a separate account from <var translate="no">TUNNEL_ACCOUNT_USERNAME</var>.
      - <var class="edit" scope="TUNNEL_SERVER_PUBLIC_IP" translate="no">TUNNEL_SERVER_PUBLIC_IP</var> with the public IP of your SSH tunnel server.

      ```bash
      ssh -N -L \
      TUNNEL_SERVER_SSH_PORT:SOURCE_DATABASE_PRIVATE_IP:SOURCE_DATABASE_PORT \
      USERNAME@TUNNEL_SERVER_PUBLIC_IP
      ```
3. At a later stage, when you [create the source connection profile](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/create-source-connection-profile.md.txt), do the following:
   1. In the **Define connection details** section, enter the private IP of your Microsoft Azure SQL Managed Instance source.
   2. In the **Define connectivity method** section, select **Forward-SSH tunnel**.
   3. Provide the public IP address or hostname of your SSH server.
   4. Provide the port that you designated for the SSH connections on the tunnel server.
   5. Enter the username for the user you created for Database Migration Service to connect as (that is, the value of <var translate="no">TUNNEL_ACCOUNT_USERNAME</var>).
   6. From the **Authentication method** drop-down menu, select the authentication method you want to use with <var translate="no">TUNNEL_ACCOUNT_USERNAME</var>:
      - If you want to use the user password, select **Password** and enter <var translate="no">TUNNEL_ACCOUNT_USERNAME</var> password in the form.
      - If you configured your SSH server to use key-based authentication, select **Private/Public key pair** and upload the **private key** [you generated in with the `ssh-keygen` command](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#generate-ssh-keys-step).

## Configure private connectivity with Private Service Connect interfaces

This connectivity method lets you connect to the private IP of your source
database without consuming
[VPC peering quota](https://docs.cloud.google.com/vpc/docs/vpc-peering.md.txt#peering-limits). Private Service Connect interfaces
are the recommended method for private IP connectivity.

To use this method,
you need to have a Cloud VPN or Cloud Interconnect configured in the same
VPC network where you intend to create the network attachment.
If you can't create the network attachment
in the VPC network where you have your Cloud VPN or Cloud Interconnect,
then you also need a reverse proxy Virtual Machine (VM) on Compute Engine to
establish the connection.

If you can't use Cloud VPN or Cloud Interconnect, we recommend
that you use the
[forward-SSH tunnel](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#forward-ssh-tunnel) or the
[IP allowlist](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#ip-allowlist)
connectivity methods instead.

To get the permissions that you need to configure Private Service Connect interfaces
ask your administrator to grant the required IAM roles on your project for the following
[accounts involved in the migration process](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/access-control.md.txt)

- User account that performs the migration:
  - To create and manage network attachments: [Compute Network Admin](https://docs.cloud.google.com/iam/docs/roles-permissions/compute.md.txt#compute.networkAdmin) (`roles/compute.networkAdmin`)
  - To create and manage private connectivity configurations: [Database Migration Admin](https://docs.cloud.google.com/iam/docs/roles-permissions/datamigration.md.txt#datamigration.admin) (`roles/datamigration.admin`)
- If you need to create a network attachment or its underlying subnetwork
  in a different project than Database Migration Service (for example, if you use a Shared VPC
  host project), then you need to grant the following role to the
  **Database Migration Service service account**
  (`service-PROJECT_NUMBER@gcp-sa-datamigration.iam.gserviceaccount.com`):
  [Compute Network Viewer](https://docs.cloud.google.com/iam/docs/roles-permissions/compute.md.txt#compute.networkViewer) (`roles/compute.networkViewer`)

For more information about granting roles, see
[Manage access](https://docs.cloud.google.com/iam/docs/granting-changing-revoking-access.md.txt).

These predefined roles contain the permissions required to perform homogeneous
SQL Server migrations with Database Migration Service. To see the exact permissions
that are required, expand the **Required permissions** section:

#### Required permissions

The following permissions are required to perform homogeneous SQL Server
migrations with Database Migration Service:

- User account that performs the migration:
  - `datamigration.*`
  - `compute.networkAttachments.create`
  - `compute.networkAttachments.get`
  - `compute.networkAttachments.list`
  - `compute.networkAttachments.update`
  - `compute.networkAttachments.delete`
- Database Migration Service service account:
  - `compute.networkAttachments.get`
  - `compute.networkAttachments.update`
  - `compute.subnetworks.get`
  - `compute.subnetworks.list`

You might also be able to get these permissions with
[custom roles](https://docs.cloud.google.com/iam/docs/creating-custom-roles.md.txt) or other
[predefined roles](https://docs.cloud.google.com/iam/docs/roles-permissions.md.txt).

To use Private Service Connect interface connectivity, do the following:

1. Set up direct connectivity with Cloud VPN to your
   Microsoft Azure SQL Managed Instance source.

   For more information, see
   [Create HA VPN connections between Google Cloud and Azure](https://docs.cloud.google.com/network-connectivity/docs/vpn/tutorials/create-ha-vpn-connections-google-cloud-azure.md.txt)
   in the Cloud VPN documentation.
2. Optional: If you can't create the network attachment in the same VPC network where you have the Cloud VPN, create [a reverse proxy Virtual Machine (VM) on Compute Engine](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-reverse-proxy.md.txt) to forward the connections between VPCs.
3. Create a network attachment in the VPC where you have your Cloud VPN. Follow these steps:
   1. In the Google Cloud console, go to the **Network attachments** page.

      [Go to Network attachments](https://console.cloud.google.com/net-services/psc/list/networkAttachments)
   2. Click **Create network attachment**.
   3. Enter a name for the attachment.
   4. From the **Network** menu, select the VPC where your source resides.
   5. For the region, use the same region where you plan to create the
      destination database.

      Database Migration Service is a fully-regional product, meaning all entities
      related to your migration (source and destination connection profiles,
      migration jobs, destination databases, conversion workspaces) must be
      saved in a single region.
   6. From the **Subnetwork** menu, select a subnet where you have at least 6 free usable IP addresses for Database Migration Service (that is, a `/29` range).
   7. In **Connection preference** select
      **Accept connections for selected projects**.

      Database Migration Service automatically adds the producer project to
      the **Accepted projects** list when you later
      [create the private connectivity configuration](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/create-private-connectivity-configuration.md.txt#create-psci-config).

      > [!CAUTION]
      > **Caution:** The **Automatically accept connections for all projects** option is less secure because it allows any service to obtain IP addresses from your subnet. We don't recommend using this option.

   8. Don't add **Accepted projects** or **Rejected projects**.
   9. Click **Create network attachment**.
4. In Database Migration Service, [create a private connectivity configuration for Private Service Connect interfaces](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/create-private-connectivity-configuration.md.txt#create-psci-config).
5. At a later stage, when you [create the source connection profile](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/create-source-connection-profile.md.txt), do the following:
   1. In the **Define connection details** section, enter the private IP of your Microsoft Azure SQL Managed Instance source.
   2. In the **Define connectivity method** section, select **Private connectivity**.
   3. From the drop-down menu, select the private connectivity configuration you created in the previous step.

## Configure private connectivity with VPC peering

To use private connectivity with Microsoft Azure SQL Managed Instance sources,
you need to have a Cloud VPN or Cloud Interconnect configured in the same
VPC network where you intend to create the private connectivity configuration
for Database Migration Service. If you can't create the private connectivity configuration
in the VPC network where you have your Cloud VPN or Cloud Interconnect,
then you also need a reverse proxy Virtual Machine (VM) on Compute Engine to
establish the connection.

If you can't use Cloud VPN or Cloud Interconnect, we recommend
that you use the
[forward-SSH tunnel](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#forward-ssh-tunnel) or the
[IP allowlist](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-azure-sqlserver.md.txt#ip-allowlist)
connectivity methods instead.

To use private connectivity with VPC peering and Cloud VPN,
follow these steps:

1. Set up direct connectivity with Cloud VPN to your
   Microsoft Azure SQL Managed Instance source.

   For more information, see
   [Create HA VPN connections between Google Cloud and Azure](https://docs.cloud.google.com/network-connectivity/docs/vpn/tutorials/create-ha-vpn-connections-google-cloud-azure.md.txt)
   in the Cloud VPN documentation.
2. Optional: If you can't create the private connectivity configuration in the same VPC network where you have the Cloud VPN, create [a reverse proxy Virtual Machine (VM) on Compute Engine](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/configure-src-connection-reverse-proxy.md.txt) to forward the connections between VPCs.
3. In Database Migration Service, [create a private connectivity configuration](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/create-private-connectivity-configuration.md.txt) to peer with the VPC network where you have your Cloud VPN.
4. At a later stage, when you [create the source connection profile](https://docs.cloud.google.com/database-migration/docs/sqlserver-to-alloydb/create-source-connection-profile.md.txt), do the following:
   1. In the **Define connection details** section, enter the private IP of your Microsoft Azure SQL Managed Instance source.
   2. In the **Define connectivity method** section, select **Private connectivity**.
   3. From the drop-down menu, select the private connectivity configuration you created in the previous step.