In order to function in a RAS Farm, a Provider (hypervisor or cloud-based) needs RAS Provider Agent to be installed in the Farm. RAS Provider Agent acts as an interface between other RAS components and a Provider. RAS Provider Agent conducts all communications with a Provider through the provider's native API.

Parallels RAS has two types of RAS Provider Agents that can be installed in a Farm:

• Built-in: This RAS Provider Agent is built into the RAS Connection Broker and is installed automatically when you install Parallels RAS. The agent can handle multiple Providers and can also be configured for high availability.

• Dedicated: This RAS Provider Agent is installed manually. It can handle only a single Provider. If you want to use this agent type with more than one provider, you need to install a separate instance for each provider.

Both built-in and dedicated RAS Provider Agents are compatible with all types of Providers supported by Parallels RAS. Which agent you choose to install depends only on your requirements. When possible, it is always recommended to use the built-in Provider Agent for high availability and business continuity.

What to read next:

• If you are adding a Provider that will use the built-in RAS Provider Agent, you may skip to Add a Provider.

• If you want to install a dedicated RAS Provider Agent on a host of your choice, read the RAS Provider Agent installation options section, which follows this one.

If you are installing a dedicated RAS Provider Agent, you first need to determine where it will be installed. Depending on the Provider type, the following options are available:

• The host on which the hypervisor is running. This option is available for Microsoft Hyper-V only.

• A supported version of Windows Server running on a physical box or in a virtual machine. For supported Windows Server versions, see Software requirements > RAS Provider Agent.

The following table lists RAS Provider Agent installation options for each supported Provider:

* High Availability is not available with these Provider Agent installation options. For details, see Enabling high availability for VDI.

In the table above, find the Provider type that you are using and see where the RAS Provider Agent can be installed. Depending on the available choices, do one of the following:

• Built-in Agent: The agent is a part of RAS Connection Broker, so it is already installed. When possible, it is always recommended to use the built-in Provider Agent for high availability and business continuity.

• Agent on a the provider: This option is only available if you are using Microsoft Hyper-V. You can simply install the agent on the host, as described in Add a Provider.

• Agent on a Windows Server (VM or HW): To use this option, make sure you have a physical box or a virtual machine running a supported version of Windows Server. You will need to specify its FQDN or IP address when adding a Provider to the Farm.

This section describes how to add a For the information on how to add a hypervisor provider, see .

Introduction

Organizations using or interested in using Microsoft Azure can provision, scale, and manage VDI and RD Session Host workloads directly from the Parallels RAS console and deploy on to Microsoft Azure using Azure Resource Manager (ARM). Parallels RAS uses a service principal with required permissions on relevant Azure resources (subscription and resource groups) to authenticate, provision and manage the resources.

Prerequisites

To use Microsoft Azure as a Provider, you need the following:

• An existing Microsoft Azure account and subscription.

• The necessary Microsoft Azure providers must be enabled, including Microsoft.ResourceGraph, Microsoft.Resources, Microsoft.Compute, Microsoft.Network.

• An ARM virtual network and subnet in your preferred region with connectivity to AD services. Microsoft Entra ID with Active Directory Domain Services (AADDS), Domain Controller in Azure IAAS or hybrid with connectivity to on-premises domain can be used.

• Site-to-site VPN or ExpressRoute is required if hybrid RAS deployment is used.

• A configured VM to be used for VDI or RD Session Host as a template.

Adding Microsoft Azure as a Provider is a two-step process:

1. First, you need to create an application in Microsoft Azure to access the resources in your subscription. This step is described in the Create a Microsoft Entra ID application section.

2. Once the application is created and registered, you can add Microsoft Azure as a Provider in the Parallels RAS Console. This step is described in Add Microsoft Azure as a Provider.

Read on to learn how to perform the steps above.

Parallels RAS VDI (Virtual Desktop Infrastructure) enables you to use host virtualization to reduce the number of physical host required to host published resources. Parallels RAS VDI supports numerous virtualization technologies, including hypervisor and cloud-based platforms.

Parallels RAS VDI also includes the Template functionality, which gives you the ability to create a template from a preconfigured host (virtual machine) and then automatically clone hosts and RD Session Host VMs from it.

When creating a template for cloning VMs in Microsoft Azure, you need to select an Azure resource group where VM clones will be created. Note that this must be a group to which you granted permissions to the Microsoft Entra ID application. You also need to select a VM size and disk type to be used for cloned VMs. These settings are specified on the Advanced page of the Create Template Wizard.

Both Virtual Desktop and RD Session Host templates can be created with Microsoft Azure as a Provider. When VMs are cloned, you will see them appear in the RAS Console. At the same time, you can also see them in the Microsoft Azure portal.

For complete information about creating and using templates, including Microsoft Azure specifics, please see the Templates section.

To complete the steps below, you must have a Microsoft Azure subscription and account. If you don't have a subscription, you need to purchase one first.

Create an Microsoft Entra ID application

An Microsoft Entra ID application is used with the role-based access control. You need to create an Microsoft Entra ID application to access resources in your subscription from Parallels RAS.

To create an Microsoft Entra ID application:

1. Log in to the Microsoft Azure portal.

2. Open the portal menu and select Microsoft Entra ID.

3. In the left pane, select App registrations.

4. Click New registration (at the top of the right pane).

5. The Register an application blade opens.

6. In the Name field, type a name you want to use for the application.

7. In the Redirect URI (optional) section, make sure that Web is selected in the drop-down list. Leave the URI field empty.

8. Click Register (at the bottom left).

9. The new Microsoft Entra ID app is created and its blade is displayed in the portal.

Note the following app properties, which are displayed at the top of the right pane:

• Display name

• Application (client) ID*

• Directory (tenant) ID*

• Object ID*

* Copy and save these properties. You will need to specify them later when adding Azure as a Provider in the RAS Console.

Create a client secret

A client secret is a string that the application uses to prove its identity when requesting a token. It essentially acts as an application password. You will need to specify this string in the RAS Console when adding Azure as a Provider.

To create a client secret:

1. If you are not on the application page anymore, navigate to it from the Home page by selecting Microsoft Entra ID > App registration and then clicking the app in the right pane.

2. In the left pane, click Certificates & secrets.

3. In the right pane, click New client secret.

4. Type a client name and select a desired expiration option.

5. Click Add. The new client secret appears in the Client secrets list.

6. IMPORTANT: Copy and save the client secret (the Value column). If you leave this page without copying the secret, it will be hidden and you will not be able to retrieve it later.

Give the application read and write access to resources

The Microsoft Entra ID app that you created must have read and write access to Azure resources. The following instructions demonstrate how to give the application read and write access to a resource group. You can also give access to a specific resource or to your entire Azure subscription. For more information, please see the Microsoft Azure documentation.

To give the app write access to the resource group where new VMs will reside:

1. In the Azure portal menu, select Resource groups.

2. Click a resource group where the new VMs will reside.

3. In the left pane, select Access control (IAM).

4. In the right pane, locate the Grant access to this resource box and click Add role assignment.

5. On the Role tab of the Add role assignment page, select Privileged administrator roles, then the Contributor role.

6. Click Next.

7. On the Members tab, select the User, group, or service principal option.

8. Click on the Select members link and enter the name of the previously created application in the Select field. Select the application in the drop-down list and click Select.

9. Click Next.

10. On the Review + assign tab, confirm that the configuration is correct and click Review + assign.

To give the app read access to the resource group:

1. Repeat steps 1-4 from the list above.

2. On the Role tab of the Add role assignment page, select Job function roles, then the Reader role.

3. Repeat steps 6-10 from the list above.

Note: If you would like to give the application read access to your entire subscription (not just a specific resource groups), select All services in the Azure portal menu, then navigate to Categories > All > Subscriptions and select your subscription. Select Access control (IAM) in the middle pane and click Add in the Add a role assignment box. Repeat steps 2-4 from the list above.

Finding your Microsoft Azure subscription ID

When you'll be adding Microsoft Azure as a Provider in the RAS Console, you will need to specify your Azure subscription ID. If you don't remember it, here's how to find it in the Microsoft Azure portal:

1. In the portal menu, choose All services.

2. In the Categories list, click All.

3. In the right pane, click Subscriptions.

4. Click a subscription and then copy and save the value from the Subscription ID field.

Summary

When you complete all of the above steps, you should have the following values saved and ready to be used to add Microsoft Azure as a Provider in the RAS Console:

• App (client) ID: Application ID.

• Directory (tenant) ID: Tenant ID.

• Client secret: Client secret (application key).

• Subscription ID: Your Microsoft Azure subscription ID.

Read on to learn how to add Microsoft Azure as a Provider in the RAS Console.

Parallels RAS supports hypervisor-based Providers and cloud-based Providers.

Hypervisors

The following hypervisors are supported:

• Microsoft Hyper-V (Windows Server 2016 up to Windows Server 2022)

• Microsoft Hyper-V Failover Cluster (Windows Server 2016 up to Windows Server 2022)

• VMware vCenter 6.5.0*, 6.7.0*, 7.x, 8.0

• VMware ESXi 6.5.0*, 6.7.0*, 7.x, 8.0

• SC//HyperCore 9.2, 9.4

• Nutanix AHV (AOS 5.15, 5.20, 6.5 LTS)

• Remote PC — This is a special type that allows you to create pools of Remote PCs. See .

* VMware ends support of vSphere 6.5.0 and 6.7.0 on October 15, 2022. While you can still use these versions with Parallels RAS 19, it is recommended to upgrade to vSphere 7.0 to ensure long-time support.

Cloud Providers

• Microsoft Azure

• Amazon Web Services

This section describes how to add a . For the information on how to add a cloud-based Provider, see .

To add a Provider:

1. In the RAS Console, navigate to Farm > Site > Providers.

2. On the Providers tab, click Tasks > Add and select the provider you want to add.

3. The Add Virtualization Provider wizard opens.

4. In the Name field, specify the name for the provider.

5. In the Description field, type an optional description.

6. In the Address field, specify the host's FQDN or IP address. For SC//HyperCore, you can specify IP addresses for several nodes.

7. Specify a user name and password to log in to the host.

8. Click the Manage Credentials button to specify the accounts that will be used to deploy RAS agents.

9. Click the Advanced Settings link to open the Advanced Provider Settings dialog. The dialog allows you choose the following options:

   • Use dedicated Provider Agent: Select this option if you will install (or have installed) the RAS Provider Agent yourself. Clear the option if you will use the .

   • Agent address: This option becomes enabled if you select the option above it. Specify the FQDN or IP address of the host where the RAS Provider Agent is (or will be) installed. This can be either a physical box or virtual machine.

   • Preferred Connection Broker: Select a RAS Connection Broker to be the preferred agent for this Provider. For more info, see .

10. Click Next.

11. The wizard will now try to connect to the RAS Provider Agent. If you specified Use dedicated Provider Agent option in the previous (optional) step, but haven't installed the agent yet, click Install and follow the instructions to push install the agent on the specified host.

    Please note that for the remote installation to work, the following requirements must be met:

    • The firewall must be configured on the host to allow push installation. Standard SMB ports (139 and 445) need to be open. See also Port reference for the list of ports used by Parallels RAS.

    • SMB access. The administrative share (\\server\c$) must be accessible. Simple file sharing must be enabled.

    • Your Parallels RAS administrator account must have permissions to perform a remote installation on the host. If it doesn't, you'll be asked to enter credentials of an account that does.

    • The target host should be joined to an AD domain.

    If push installation cannot be performed for any reason, you can install the agent manually using the installer. See .

12. If you've selected Microsoft Hyper-V Failover Cluster as the Provider type, the page opens where you can disable MAC address management for hosts. Note that you should only do it if you are using Microsoft System Center Virtual Machine Manager (SCVMM) or other solution to manage MAC addresses. See the explanation below.

    MAC address management is required when using Microsoft Hyper-V Failover Cluster as a Provider. This is to avoid duplicate MAC addresses, which may occur when a host is migrated to a different node in the cluster and the MAC address is released and reused on the original node. If that happens, such a host can no longer be managed in a Farm. Parallels RAS uses a pool of static MAC addresses at the Provider level to automatically generate and assign MAC addresses to hosts. This way, when a host is migrated to a different node in the cluster, its MAC address will not be reused for a different VM and no duplicate MAC addresses will occur. The pool has 10,000 reserved MAC addresses in the range displayed in the Starting MAC address and Ending MAC address fields on the wizard page.

    As was said above, if you are already managing MAC addresses using SCVMM or other solution, clear the Enable MAC address management option.

13. Click Next.

14. If you've selected VMware vCenter as the Provider, another page opens (the page will not open for any other host type). On this page, you can specify a vCenter resource pool. This allows you to enumerate VMs by selecting a cluster (root resource pool) or an individual resource pool within a cluster. To choose a resource pool, select the Use specific resource pool option and then click the [...] button next to the Resource Pool field. In the dialog that opens, select a desired resource pool. Note that if you leave the Use specific resource pool option cleared, all VMs from the entire vCenter cluster will be retrieved (max number is 35,000). Click OK when done.

15. Click Finish to close the wizard.

To create the IAM user account, you can use the AWS Management Console, the AWS CLI,?Tools for Windows PowerShell, or AWS API operation. In this example, we will be using the AWS Management Console:

1. Sign in to the AWS Management Console and open the IAM page at .

2. In the navigation pane, choose Users and then click the Add users button.

3. Under Set user details section, provide a user name such as "ParallelsConnector".

4. Under AWS access type, select Access key - Programmatic access, as the Parallels RAS Console will be using APIs to communicate with your AWS account. This will create an access key for the IAM user. You can view or download the access keys when you get to the Final page. Click Next to proceed to the permissions page.

5. On the permissions page, you can create a user group for the new IAM user to be a part of. This is recommended as its beneficial for management purposes, although not mandatory.

6. If you are not using groups, choose Attach existing policies directly. A list of the AWS managed and customer managed policies in your account will appear.

7. Filter policies and choose AmazonEC2FullAccess, which is an AWS managed preconfigured policy, and click Next to proceed to the next page.

8. Optionally, on this page, you can use the tags to organize, track, or control access for this user.

9. Once the tags are ready, click Next to see all of the choices you made up to this point. When you are ready to proceed, click Create user.

10. To view the user's access key ID and secret access keys, click Show next to each password and access key that you want to see. To save the access keys, choose Download CSV and then save the file to a safe location.

    Please note that this is your only opportunity to view or download the secret access keys.

11. Save the user's new access key ID and secret access key in a safe and secure place to be used next in Parallels RAS Console.

Proceed to

To add Microsoft Azure as a Provider:

1. In the RAS Console, navigate to Farm > Site > Providers.

2. On the Providers tab, click Tasks > Add > Microsoft Azure.

3. The Add Cloud Computing wizard opens.

4. In the wizard, specify the following:

   • Name: Name of the provider.

   • Description: Description of the provider.

   • Manage credentials: the administrative accounts that will be used to deploy Parallels Agents.

   • Authentication URL: Prepopulated with the Microsoft authentication site URL. Unless otherwise required or indicated, keep the default value provided.

   • Management URL: Prepopulated with the Microsoft Azure management site URL. Unless otherwise required or indicated, keep the default value provided.

   • Resource URI: Prepopulated with the Microsoft Azure resource URI. Unless otherwise required or indicated, keep the default value provided.

   • Tenant ID: The "Directory (tenant) ID" value of the Microsoft Entra ID app that you created earlier.

   • Subscription ID: Your Microsoft subscription ID.

   • Application ID: The "App (client) ID" value of the Microsoft Entra ID app that you created earlier.

   • Application key: The "Client secret" value of the Microsoft Entra ID app that you created earlier.

5. Click the Advanced Settings link to open a dialog where you can configure the following optional settings:

   • Use dedicated Provider Agent: When this option is cleared (default), the built-in RAS Provider Agent will be used. If you want to use a dedicated RAS Provider Agent, select this option and specify the host FQDN or IP address.

   • Agent address: This option becomes enabled if you select the option above it. Specify the FQDN or IP address of the host where the RAS Provider Agent is (or will be) installed. This can be either a physical box or virtual machine.

   • Preferred Connection Broker: Select a RAS Connection Broker to be the preferred agent for this Provider. For more info, see Enabling high availability for VDI.

6. Click Next. The wizard will display the new Provider information and will indicate the RAS Provider Agent status. If everything is OK, click Finish to exit the wizard. If something is not as expected, click Back and correct any mistakes if necessary.

The new Provider will now appear on the Providers tab in the RAS Console. Complete the Provider addition as follows:

1. Click Apply to apply the changes.

2. Verify the value of the Status column. If it's anything other than OK, right-click the Provider and choose Troubleshooting > Check agent. Verify the agent status and install it if necessary, then click OK. The Status column on the Providers tab should now say OK.

Modifying the Provider configuration

To view and modify the Provider configuration, right-click it and choose Properties. In the dialog that opens, view and modify the Provider properties.

Read this section to learn how manage VDI components in Parallels RAS.

By default, Provider Agent is installed together with Connection Broker. However, if you want to install Provider Agent on a separate server or the push installation from the RAS Console cannot be performed for any reason. If that happens, you can install the agent by running the installer directly on the target server.

To install the dedicated Provider Agent Agent:

1. Log in to the server where you want Provider Agent installed using an administrator account and close all other applications.

Copy the standard Parallels RAS installer (RASInstaller.msi) to the server and run it:

1. When you get to the Select Installation Type page, select Custom and click Next.

2. Click on RAS Provider Agent dedicated and select Entire Feature will be installed on local hard drive from the drop-down list.

3. Ensure that all other components are cleared (excluded from the installation) and click Next.

4. Click Install and follow he onscreen instruction to install the agent.

The Provider Agent does not require any configuration. Once it is installed, go back to the RAS Console, highlight the server name and click Troubleshooting > Check Agent. If the agent is installed properly, the status should change to Agent Installed.

To uninstall the Provider Agent from a server:

1. Navigate to Start > Control Panel > Programs > Uninstall a Program.

2. Find Parallels Remote Application Server in the list of installed programs.

3. If you don't have any other Parallels RAS components on the server that you want to keep, right-click Parallels Remote Application Server and then click Uninstall. Follow the instructions to uninstall the program. You may skip the rest of these instructions.

4. If you have other RAS components that you want to keep on the server, right-click Parallels Remote Application Server and then click Change.

5. Click Next on the Welcome page.

6. On the Change, repair, or remove page, select Change.

7. On the next page, select Custom.

8. Select RAS Provider Agent dedicated, then click the drop-down list in front of it, and click Entire feature will be unavailable.

9. Click Next and complete the wizard.

To verify that the RAS Provider Agent is installed and functions properly, do the following:

1. First, you can look at the Status column in the Farm > Site > Providers list. If there's a problem with the agent, the column will display an appropriate description. Note that in addition to the description, the Status column uses a color code to indicate the agent status as follows:

   • Red — Not Verified

   • Orange — Needs Update

   • Green — Verified

2. Right-click a host and then click Troubleshooting > Check agent in the context menu.

3. The Provider Agent Information dialog opens displaying the information about the Provider Agent, VDI Services, and other related info.

4. If the Provider Agent is not installed, click the Install button and follow the onscreen instructions. See RAS Provider Agent installation options for more info.

Read this section to learn how to modify the configuration of a Provider in Parallels RAS.

Configure a Provider

To configure an existing Provider:

1. In the RAS Console, navigate to Farm > Site > Providers. .

2. In the Providers tab, select a Provider and click Tasks > Properties. The Properties dialog opens.

Enable or disable a Provider in Site

By default a Provider is enabled. To enable or disable a Provider, use the Enable provider in site option on the General tab.

Properties: configure Provider connection settings

The General tab has different properties depending on whether it's a hypervisor-based or cloud-based provider.

Hypervisor Provider:

• Type: Provider type.

• Subtype: Hypervisor version. If the hypervisor version that you are using is not listed, select Other.

• Host: The Provider host IP address.

• Port: Port number on which the Provider listens for incoming connections.

• Resource pool: This field is enabled for VMware vCenter only. If you've specified a vCenter resource pool while adding a Provider, the pool will be displayed here. The [...] button allows you to specify a different pool (or select one if the field is empty), but only if no hosts from the current pool have been created or used in Parallels RAS in any way. If Parallels RAS detects any current usage, you will see a warning message and will not be able to change it. If you still want to select a different resource pool, you'll have manually do a full clean up in the RAS Console, so that no usage of any kind exists.

• Description: An optional description.

• Dedicated Provider Agent: Select this option if you have a dedicated RAS Provider Agent installed on a different host. Enter the host FQDN or IP address in the Agent address field.

Cloud-based Provider:

• Type: Cloud-based Provider type (e.g. Microsoft Azure).

• Name: Provider name.

• Description: An optional description.

• Credentials: Credentials for the account used for installing RAS Guest Agent.

• Dedicated Provider Agent: Select this option if you have a dedicated RAS Provider Agent installed on a different host. Enter the host FQDN or IP address in the Agent address field.

For description of the remaining properties of Microsoft Azure, please see Add Microsoft Azure as a Provider.

Credentials: configure username and password

The Credentials tab has different properties depending on whether it's a hypervisor-based or cloud-based host.

Hypervisor Provider:

• Specify the username and password to log in to the Provider. Click the Check Credentials button to verify the credentials that you've entered.

Cloud-based Provider:

• See Add Microsoft Azure as a Provider or Adding AWS as a Provider.

Advanced

The Advanced tab allows you to configure a provider to automatically change the type of the used managed disk to Standard HDD for VMs that are not currently in use. When a VM is started, the managed disk is automatically changed to the original type. This feature allows you to reduce the cost of maintaining VMs.

To enable disk storage cost optimization:

1. Right-click a provider in the list and choose Properties.

2. In the provider properties window, select the Advanced tab.

3. Select the Enable disk storage cost optimization option.

4. Select the desired option in the Set timeout before enabling storage cost optimization drop-down list.

MAC addresses

This tab is only displayed for Microsoft Hyper-V Failover Cluster as the Provider. It is used to enable or disable MAC address management for hosts. For more information, please see Add a hypervisor host (read the description of the step where the MAC address management is configured).

Please note that MAC address management is available in Parallels RAS since version 18. In new Parallels RAS 18 installations, this functionality is enabled by default when a Provider of type Microsoft Hyper-V Failover Cluster is added to a farm. In older Parallels RAS versions, the functionality is disabled for existing Providers, but is enabled by default when a new Provider is added.

To configure Amazon Web Services as a Cloud Computing provider:

1. In the RAS Console, navigate to Farm > Providers.

2. Click the Tasks drop-down menu and choose Add (or click the [+] icon).

3. In the menu, select Amazon EC2. The Add Cloud Computing Provider wizard opens.

4. In the Wizard, specify the following:

   • Name: Name of the provider.

   • Description: Description of the provider.

   • Manage credentials: the administrative accounts that will be used to deploy Parallels Agents on the session hosts (Amazon EC2 instances). The current RAS administrator is already present in this list, but you can other accounts.

   • Access Key ID: Your access key ID.

   • Secret Access Key: Your secret key.

5. Click Next.

6. Wait until Parallels RAS validates the settings and click Next.

7. Select the Region that you will use. In most cases, the best Region would be the one closest to you. You can also choose one of opt-in AWS Regions by selecting the Opted-in Region option or specify a custom EC2 endpoint URL by selecting the EC2 Endpoint URL option.

8. Click Finish.

9. Proceed to creating a Template as described in Creating a VDI template. During template creation you can configure the instance type for the clones and the storage including Type, Size, and IOPS. Note that you can also do this from Farm > RD Session Hosts > right-click the template > Properties.

Introduction

Amazon Web Services (AWS) is a leading cloud platform provider offering over 200 fully featured services from data centers globally. Parallels RAS 19 provides the ability to integrate, configure, maintain, support, and access Amazon EC2 workloads on top of the existing capabilities of Parallels RAS.

Support is targeted at multi-session (RDSH), single session (server-based VDI) server operating systems, and other Microsoft operating systems, if your organization holds licenses for them. For more information about using Microsoft operating systems with AWS, see https://aws.amazon.com/windows/faq/.

Parallels RAS Console allows you to do the following:

• Manage Amazon EC2 instances

• Create and manage templates

• Create and manage instance pools

• Configure autoscaling

• Enable, reboot, start up and shut down instances via schedules

• Configure image optimization

• Use FSLogix Profile Container and MSIX app attach

• Change instance types and storage types

Prerequisites

• An AWS account. If you do not already have an account, you can create it for free at aws.amazon.com/ec2/.

• A working Microsoft Active Directory environment to join the Amazon EC2 cloned instances to your domain.

• A preconfigured Virtual Private Cloud (VPC) as your virtual network and security groups that act as a virtual firewall for your EC2 instances.

• A preconfigured Amazon EC2 instance, which will be used later as a Parallels RAS template, running on Windows Server 2012 up to Windows Server 2022.

Use the Wildcard input field at the bottom of the Pools tab to specify a wildcard to indicate which hosts should be available for users. If a VM name matches the wildcard, it will be available. If not, the users will not be able to use it. Use the asterisk operator (*) to specify a wildcard (e.g. ABC*, *ABC*).

This topic describes how to use the same Provider in multiple RAS Farms simultaneously. To better understand the problem and the solution, consider the following hypothetical example:

• Let's say we have a hypervisor with two available virtual machines.

• We also have two Farms (1 and 2).

• Our intention is to use the first host to host resources in Farm 1 and the other to be a template in Farm 2. Both hosts will run simultaneously on the hypervisor, but each one will be available in its respective Farm only.

The problem is, RAS Guest Agent can normally communicate with one RAS Provider Agent, but since each Farm has its own RAS Provider Agent, this will not work out of the box. The solution is to make the RAS Guest Agent running in a host to be aware of only one specific RAS Provider Agent with the ability to change the assignment as needed.

The assignment is done via Windows registry. All hosts belonging to VDI pools and host clones created from a template need to have a new String value 2XVDIAgent specifying the RAS Provider Agent name or address. To add the value:

1. Log in to Windows running in the virtual machine, open the registry editor (regedit) and locate the following keys:

   • 32-bit systems: HLKM\Software\Parallels\GuestAgent

   • 64-bit systems: HLKM\Software\WOW6432Node\Parallels\GuestAgent

2. Add a String value named 2XVDIAgent. The value data should be specified as follows:

   • If a dedicated RAS Provider Agent is used, the value must be set to the FQDN or IP address of the server where the agent is installed.

   • If the built-in RAS Provider Agent is used with manual agent selection, the value must be set to the FQDN or IP address of the RAS Connection Broker.

   • If the built-in RAS Provider Agent is used and the agent is selected automatically (high availability), the string must contain FQDNs or IP addresses of all RAS Connection Brokers separated by a semicolon (i.e. <PA1 address>;<PA2 address>;<PA3 address>).

Note that you can include names or IP addresses of multiple Connection Brokers for the manual agent selection scenario as well (the second bullet in the list above). This way you will not need to change the value every time you switch the preferred Connection Broker for a Provider.

Pools offer administrators more flexibility when managing an extensive number of hosts, especially when they are implemented in large company infrastructures. The RAS Console provides you with the framework and tools needed to create a complete pool management foundation. To manage pools, in the RAS Console, navigate to Farm > <Site> > VDI and then click the Pools tab.

To delete a host pool:

Right-click it and then click Delete (or click the minus-sign icon, or Tasks > Delete).

This section contains design advice that you might want to keep in mind when using AWS in Parallels RAS.

DHCP options set

You might need to use an AWS DHCP options set to specify a custom DNS pointing to the domain controller so that the VMs created from templates are able to join the Active Directory domain. If the custom DNS is not set, the default AWS public DNS will be used, and the VMs won't be able to communicate with the domain controller.

For information on how to configure DHCP options sets, see https://docs.aws.amazon.com/vpc/latest/userguide/DHCPOptionSet.html.

The Provider Agent and Guest Agents need to be on the same subnet for the Guest Agent to discover the Provider Agent using broadcasts. If this is not possible, then a registry setting with the IP of the Provider Agent needs to be added on the VM as described here: https://kb.parallels.com/en/124157?language=en.

Service quotas

Sometimes solutions scale in usage, invocations, number of instances, and so on. Due to this, the standard AWS service quotas can be reached. For more information about AWS service quotas, see https://docs.aws.amazon.com/general/latest/gr/aws_service_limits.html.

Parallels RAS integrations are subject to the EC2 and EBS endpoint limits as specified here:

• https://docs.aws.amazon.com/general/latest/gr/ec2-service.html

• https://docs.aws.amazon.com/general/latest/gr/ebs-service.html

Encrypting storage

The storage of clones created from RAS templates will be encrypted if the AWS administrator enables encryption of the RAS template VM storage in AWS Management Console.

Encryption can be enabled by default or explicitly when launching a new EC2 VM:

For more information about encryption, see https://aws.amazon.com/blogs/compute/must-know-best-practices-for-amazon-ebs-encryption/.

To add a host pool:

1. In the RAS console, navigate to Farm > <Site> > VDIHost pools.

2. Click the Tasks drop-down list above the Host Pools list and then click Add (or click the plus-sign icon). This opens the Add VDI host pool wizard.

3. Select Enable Host pool in site to enable the host pool. Specify the name and the description for the new host pool.

4. Click Next.

5. On the Provisioning page, select whether this host pool will contain template-based or standalone hosts:

   • Template: Hosts will be created dynamically from a template. You will need to create or select an existing template in the next step or later. Choosing Template as the provisioning type ensures a homogeneous host pool, which is recommended to provide consistent user experience across the host pool.

   • Standalone: Select one or more hosts that already exist. You'll be able to do it in the next step or you can do it later. Prior to adding hosts to host pools, ensure that hosts are domain joined and have network access to the domain environment. Note that the Standalone provisioning is considered "unmanaged" as it lacks some of the functionality, such as Autoscaling.

6. Depending on the selection made on the Provisioning page (above), do one of the following

   • Standalone: Select one or more hosts from the list to be included in the host pool (you can also add hosts to the pool later).

   • Template: Select a template from the list or click Create new to create a new template and specify the template settings. Versions: If you selected an existing template, select one of its versions.

7. Click Next.

8. (Templates only) On the Provisioning page, specify the following options:

   • Template name: Choose and type a template name.

   • Host name: A pattern to use when naming new hosts.

   • Number of hosts deployed on wizard completion: The number of hosts to deploy once the template is created. Please keep in mind that this will take some time because the hosts will be created one at a time.

   • Host state after the preparation: Select the power state that should be applied to a host after it is prepared. Choose from Powered on, Powered off, or Suspended. Note that when the power state is set to Power off or Suspended, the number of running (fully ready and waiting for incoming connections) hosts is controlled by the Keep available buffer setting (see above). For example, let's say the Maximum hosts value is set at 200, the number of guest hosts deployed on wizard completion is 100, and the power state after preparation is Powered off. The result of such a configuration will be 100 clones deployed and powered off.

   • Maximum hosts: Specify the maximum number of hosts that can be created from this template.

   • (Only for templates created from AWS instances) Virtual networks: Select Availability Zones for the host pool. Choose from Use the subnet associated with the template and Use specific subnets available in the VPC. If you select Use specific subnets available in the VPC, then after clicking Next you can manaully select one or several Availability Zones.

   • Autoscale settings: Configure autoscale settings as described in .

9. Click Next.

10. On the Host pool settings page, specify the following options:

    • On session: Select when an action triggers.

    • Perform action: Select an action.

    • After: Select how much time must pass before action triggers.

11. Click Next.

12. On the User profile page, you can select from Do not manage by RAS (user profiles will not be managed) or FSlogix. Microsoft FSLogix Profile Container allows to maintain user context in non-persistent environments, minimize sign-in times and provides native profile experience eliminating compatibility issues. For complete instructions, please see .

13. Click Next.

14. (Standalone only) On the Optimization page, configure optimization as described in .

15. On the Summary page, review the template summary information. You can click the Back button to correct some of the information if needed.

16. Finally, click Finish to create the host pool and close the wizard.

A VDI pool can contain different types of members. These could be all available hosts, specific hosts, and hosts created from a template.

Adding a member to a host pool

To add a member to a pool:

1. Double-click a pool in the Host pools list.

2. Select the Members tab.

3. Click the plus (+) button and choose a member type from the following list:

   • All hosts in provider. All hosts that are located on a given Provider. After clicking this options, you'll be able to select a Provider.

     Note: Parallels does not recommend to use this type because there's a possibility that hosts with unsupported OS will be added (e.g. Linux, HALB etc). If you need to use this type, please do it carefully or use a .

   • Host. A specific host located in the Farm. After clicking this options, you'll be able to select a host from the list.

   • Resource pool. A group of hosts that were natively configured in the hypervisor as a pool. Please note that a hypervisor may use a different term for pools (e.g. "resource pools"). After clicking this option, you'll be able to select a resource pool from the list, if any are available.

   • Template. Hosts that are automatically created from a template. After selecting this option, you'll be able to select a template. For more information about templates, refer to .

4. After you click one of the above menu items, you will be presented with the list of the available hosts, hosts, or templates from which you can make your selection.

Removing a member from a host pool

To remove a member from a pool:

1. Double-click a pool in the Host pools list.

2. Select the Members tab.

3. Select the pool member that you wish to delete.

4. Click the minus (-) button.

When a member is removed from a host pool, it is deleted.

You can enable and configure automatic updates for all VDI hosts in a host pool.

Schedule Agent auto-upgrade

To schedule Agent auto-upgrade:

1. Go to Farm > Site > VDI > Host pools > Properties > Auto-upgrade tab.

2. Clear the Inherit default settings options if you want to modify them for this host pool.

3. Select the Enable auto-upgrade maintenance window option. During the maintenance window, all hosts in the host pool will try to download Agent upgrades. The upgrades will be downloaded and installed as soon as all users log out of their hosts. New logons from users are prohibited (drain mode). If the users don't log off during a maintenance window, the upgrades won't be installed until the next window.

4. Specify the start date and time, duration, and recurrence settings for this event. To make this a one-time event, select Never in the Recur drop-down list.

5. (Optional) If you want to forcefully log off all users and download the upgrades at the end of a maintenance window, select the Force logoff of current sessions at the end of the maintenance window duration option.

6. (Optional) Configure a message that will be sent to users before or during a maintenance window. Click the Configure messages button and specify the message title, body, and the time period when it should be sent.

Cancel Agent auto-update

To cancel Agent auto-update:

1. Go to Farm > Site > VDI > Host pools.

2. Select Tasks > Cancel auto-upgrade maintenance window.

Virtual desktop templates are the essential part of the Parallels RAS VDI. They are used to create hosts for publishing of desktops, applications, documents, etc. The guest OS support is the same as for RAS Guest Agent (which must be installed in a VM). See Software requirements.

Hosts created from a Virtual desktop template normally serve a single user. They are managed entirely from within the RAS VDI, which includes such features as creating persistent VMs, managing VDI sessions, publishing resources from a specific Virtual desktop template, and others.

When you create a template, it is normally managed by a single Provider, which is the same provider to which the source virtual machine belongs. Clones are deployed from the template and run on the same Provider, which was used to create the template. Cloned hosts are usually stored on a centrally shared storage, such as Storage Area Network (SAN).

Beginning with Parallels RAS 18, admins have the ability to create a template and distribute it to multiple Microsoft Hyper-V hosts. As a result, the template configuration, as seen in the Parallels RAS Console, is shared among multiple Microsoft Hyper-V hosts, while each host has its own copy of the template residing on its local storage. This makes it possible to deploy clones not only to a centrally shared storage, but also to local disks of multiple independent Microsoft Hyper-V hosts. Scaling out is easily carried out by adding as many Microsoft Hyper-V hosts to the template distribution list as necessary.

Template distribution is configured on the Distribution page of the Create Parallels Template Wizard, which is described in the subsequent sections. If you are planning on using the template distribution functionality, please read the Prerequisites subsection below before running the wizard.

For more information about managing multi-provider distribution for a template, see also Managing multi-provider template distribution.

Prerequisites

• Template distribution is supported on a standalone Microsoft Hyper-V Server 2012R, 2016, 2019, 2022.

• All target Providers must have identical:

  - Provider type and subtype.

  - Folder path where hosts will reside.

  - Virtual switch name to which hosts are connected.

• Hyper-V hosts must be domain-joined. The current implementation uses a full VM copy of the template to distribute the template to other hosts (local storage) via the Hyper-V Live Migration mechanism.

• The Microsoft Hyper-V server hosting the source VM may also be used as a target host.

• Always ensure that enough storage space is available prior to choosing target hosts to which the template will be distributed and on which clones will be created.

• Hyper-V settings must have Kerberos authentication enabled and appropriate delegations configured in AD:

  - Go to Hyper-V settings for the host machines and enable Live Migration using Kerberos.

  - Go to Active Directory Users and Computers and for each Hyper-V host server enable delegation for "cifs" and "Microsoft Virtual System Migration Service" for all servers you want to migrate To and From.

Requirements

To complete the tasks described in this section, the following requirements must be met:

• For hypervisor-based hosts, make sure the hypervisor tools are installed and running in the host.

• Make sure you know account credentials that will allow you to push install the agent software on a VM. If you run the Parallels RAS console using such credentials (e.g. a domain admin), you will not be asked to enter them during the agent installation. If you run the console using a different account, you'll be asked to enter credentials when you install the agent.

• The guest OS (Windows) running in the VM must be configured to obtain an IP address from a DHCP server.

• For users to access published resources in a host, the RDP port must be open locally or via Group Policy in Windows running in the VM. The default RDP port is 3389.

• For RD Session Host templates, Network Discovery UDP port 137 must be enabled for a domain firewall profile in the guest OS. This can be done via domain group policies or manually in the guest OS.

Manual agent installation

Normally, you will push install the necessary agent software in a source VM right from the Parallels RAS console (as described later in this section). However, you can also install the software manually by running the Parallels RAS installer in Windows in the VM. When doing so, use the Custom installation option and select RAS Guest Agent to be installed in the source VM.

Create a template

To begin creating a template:

1. In the RAS Console, navigate to Farm > <Site> > VDI.

2. Select the Templates tab in the right pane.

3. In the Tasks drop-down list, click Add (or click the "+" icon)

4. In the dialog that opens, select a host from which you would like to create a template and click OK.

5. The Create Parallels Template Wizard opens. Each wizard page is described below in the order they appear on the screen.

6. Verify that the Agent is installed and install it manually if needed as described in Step 1: Check and install the Agent. This step only appears if an on-premises Provider is used.

7. Configure the template as described in Step 2: Configure the template.

Hosts that belong to a pool (and other hosts and desktops) are managed on the VDI > Desktops tab, where you can perform all of the standard desktop management operations from the Tasks menu. The operations include Recreate, Delete, Upgrade all Agents, Assign, Unassign, Show sessions, Start, Stop, Restart, Suspend, Reset, and others. The Restart operation (graceful) has a 10 min timeout. If not completed during this time, the Reset operation (forced) will be used.

By default, the Desktops tab displays all of the desktop available in the Farm (you may need to scroll the list to see all available desktops). To see just the hosts that belong to a specific pool, select a pool in the Pools tab and click Tasks > Show hosts in Pool. This will switch you to the Desktops tab where the list will be automatically filtered to include only the VMs that belong to the selected pool.

Templates are used to automate the creation and deployment of hosts in Parallels RAS. A template is based on an existing virtual machine created with one of the hypervisors supported by Parallels RAS. Once a template is ready, it can be used to create clones (hosts) that will inherit all properties of the template. The resulting hosts can then be used to host published resources.

Read the following topics to learn how to create and use a template.

The Advanced page has different properties for different types of Providers. The differences are described below.

Hypervisor-based Providers:

• Cluster Shared Volume (CSV), Network share: These two options appear if you are using Hyper-V Failover Cluster. They allow you to select a type of storage where hosts will be created. Select a desired option and then click the [...] button next to the edit field. Depending on the option selected, specify a Cluster Shared Volume or network folder. Note that a shared folder must be compatible with SMB 3.0. Please also note that the same credentials used to register Microsoft Hyper-V host as a Provider will be used to access the SMB file share for Hosts.

  Please also read the important note below.

• Folder: This option is available if you are using Hyper-V, VMware vCenter, or Nutanix AHV (AOS). It specifies a folder where hosts will be created.

• Use a separate network interface for LAN access: This option is available if you are using any of the Hyper-V or VMware providers. Specifies the network interface that will be used by Connection Broker and Provider Agent. This is useful if a template has several network interfaces and you want to use a specific one for communication with Parallels RAS. If you select this option, you also need to specify the following:

  • Address: IP address of the network interface.

  • Subnet mask: Subnet mask of the IP address.

• Resource pool: Specifies a VMware resource pool.

• Physical Host: Available for VMware vCenter. Specifies a physical host where hosts will be created.

• Enable hardware acceleration graphics licensing support: This option is available if you are using VMware vCenter or VMware ESXi. Select it to allow vGPU-enabled hosts to unregister their vGPU licenses from the license server on shutdown.

Microsoft Azure Provider:

• Resource group: Select an Azure resource group where the cloned VM will be created. Note that this must be a group to which you granted permissions to the Microsoft Entra ID app. For details, see .

• Size: Select a VM size to be used for cloned VMs.

• OS disk type: Select a disk type to be used for cloned VMs.

This page is used to configure template distribution to multiple Microsoft Hyper-V hosts. Note that this page will only appear if the source VM is a Microsoft Hyper-V machine. For the description of this feature and requirements, please see .

To configure template distribution:

1. Select the Enable multi-provider template distribution option.

2. In the Available list, select one or more providers and click Add (or Add all to add all available providers). Note that only providers of the same type and subtype as the source VM are displayed in this list.

3. In the Number of providers for concurrent distribution field, specify the number of concurrent distribution operations. The template is distributed to target hosts using Hyper-V Live Migration, which first exports the virtual machine to a file and then moves it to the destination host. For each host in the Target list, a Live Migration operation must be performed. The number specified here dictates how many network copy operations should be started at the same time. The larger the number, the more network resources will be required. Note that virtual machine exports (the first step of Live Migration) are always done one VM at a time, so the number you specify here affect only the copy operations.

When done, click Next to proceed to the next wizard page.

On the License Keys page, specify the license key information that will be used to activate virtual machines created from this template.

First, select the license key management type that you are using in your organization (KMS or MAK). Parallels recommend to use KMS because MAK has limited activations.

Key Management Service (KMS): If you are using KMS, click the Finish button to save the template configuration information. Virtual machines that will be created from this template will look for KMS in DNS (at the end of the OS mini-setup and domain joining) and will be activated accordingly.

Multiple Activation Keys (MAK): If you are using MAK, do the following:

1. Click the Add button and type a valid key in the License key field.

2. In the Max. guests field, specify the key limit. The limit should be greater than or equal to the max guests in the template (which you set on the first page of the wizard)

3. Click OK.

On the Summary page, review the template summary information. You can click the Back button to correct some of the information if needed.

Finally, click Finish to create the template and close the wizard.

The Optimization page allows you to specify settings that will be used to optimize Windows running in hosts for best performance in a Parallels RAS environment. You can select Windows components, services, and other options that will be disabled, removed, or optimized to ensure a more efficient, streamlined, and improved delivery of virtual apps and desktops. For the complete description, please see Optimization.

After reading the Optimization section mentioned above, please also note the following VDI specifics:

• Optimization is disabled by default when you create a new template. If you plan to enable it, you should make a backup (create a full clone) of the source host before doing so. You can also create a template with optimization disabled, then create a snapshot, and only then enable optimization. Making a backup is a good idea because once optimization settings are applied, they cannot be rolled back.

• To enable optimization for an existing template, the template must be in maintenance. A template in the "Ready" state has the Optimization tab disabled.

• When optimization is enabled or modified for an existing template and the template exits the maintenance mode, the administrator will be asked to recreate existing hosts, so that optimization settings are applied to them. Note that hosts must be recreated (now or later) to receive optimization settings.

• When optimization is applied to a template, its status changes to Optimization in progress (among others). At this stage, you can select the template in the list and click Tasks > Stop optimization, which will cancel the operation.

On the General page, specify the following options:

• Template name: Choose and type a template name.

• Clone method: Whether to create linked or full clones. A full clone is a complete copy of a template. As such, it occupies as much space on the physical hard drive as the source template and takes a significant time to create. A linked clone is a copy of a template made from a snapshot that shares virtual disk with the source template, therefore it takes much less space on the physical hard drive and it takes only a couple of minutes to create.

  You should use full clones if your application and OS updates are too slow (full clones take longer to create, but they provide the best possible performance). Otherwise if your updates are fast enough, use linked clones as it takes much less time to create them.

• (Microsoft Azure only) Availability set: Select a Microsoft Azure availability set.

This step only appears if an on-premises Provider is used. It will not appear for Azure Virtual Desktop and cloud providers.

In this step, the wizard will check if the selected VM has the RAS Guest Agent installed. Wait for it to finish and then examine the Status field (closer to the bottom of the page). Depending on the result, do one of the following:

• If the agent is installed, click Next to continue. You may stop reading here and jump to Step 2: Configure the template.

• If the agent is not installed, you need to install it as described below.

To install the agent, first click the Customize Guest Agent deployment settings link and specify the options in the dialog that opens. None of the options are forced, so you can select or clear them depending on your needs. Note that depending on the template type, the options are different, as described below.

Virtual desktop:

• Add firewall rules: Automatically configure firewall rules in the host.

• Allow remote desktop connections: Select to automatically configure remote desktop access in the VM.

• Specify users or groups to be added to the Remote Desktop Users group: Select this option and then click the [+] icon to add specific users to the group.

RD Session Host:

• Add firewall rules: Automatically configure firewall rules in the host.

• Note: Network Discovery UDP port 137 must be enabled for a domain firewall profile in the guest OS as a separate step. This can be done via domain group policies or manually in the guest OS.

• Install RDS role: Install the RDS role in the host.

• Enable Desktop Experience: Enable the Desktop Experience feature in Windows.

• Restart server if required: Restart the VM if required.

• Specify users or groups to be added to the Remote Desktop Users group: Select this option and then click the [+] icon to add specific users to the group.

When done specifying the options, click OK to close the dialog.

Now click the Install button and follow the onscreen instructions to install the agent software.

Once done, verify that the agent is installed by looking at the Status field on the Check Agent wizard page. If so, continue to the next section that describes Step 2: Configure the template.

After a template is created, Parallels RAS begins creating hosts from it, one virtual machine at a time. The number of VMs created at this time is determined by the Number of hosts deployed on the wizard completion property (all property names here and later refer to the Create Template Wizard described earlier).

The number of VMs available at any time will never go below the number specified in the Keep available buffer property. To comply with this rule, a new VM is automatically created when needed. At the same time, the total number of VMs will never exceed the number specified in the Maximum hosts property.

Please note that creating a new host from a template takes some time, especially when a template is configured to create full clones (linked clones are created much faster). If a host is in the middle of being created, and no other VMs are available, a user (or users) who need it will have to wait until the VM is ready.

If a host encounters a problem during the preparation stage, it will remain on the server in unusable state. You can identify such VMs by the Failed to create value in the Status column. Unless a VM like this is repaired or recreated, it will be automatically removed after the time period specified in the Auto remove hosts which failed preparation after field in Site defaults (Farm > <Site> > VDI > Desktops > Tasks > Site defaults). You can view the details of the failure to create a host by clicking on the Details link in the Status column or double-clicking the column. You can also choose Recreate or Delete from the same menu. For more information on how to recreate a host, please see the Template maintenance section.

Auto-deletion of hosts

A host is automatically deleted when it is not used longer than specified in the Delete unused hosts after field in template properties.

When you create VDI host pools, you can assign a template to a pool. This can be done when you create or modify a pool, or it can be done from the Templates tab.

To assign a template to a host pool:

1. On the Templates tab, select a template.

2. Click Tasks > Assign to pool. A wizard opens.

3. On the Versions page, select the template version that will be assigned to the host pool.

4. (Optional) On the Host pool page, select the host pools that you want to recreate on schedule and click the Next button. You will see a dialog that allows you to schedule recreation. Configure the schedule according to your needs and click Next.

5. Click Finish.

To remove a template from a host pool:

1. Select a template and click Tasks > Remove from pool.

2. A dialog opens listing all host pools to which this template is assigned.

3. Select the host pools to remove the template from and click OK.

Note that if a host pool has hosts created from the template that you are removing, they will be removed as well. A message is displayed where you need to confirm the removal.

Hosts are created from a template automatically. In a situation when one or more additional hosts are required, you can add (create) them manually.

To add a host:

1. In the RAS Console, navigate to Farm > <Site> > VDI > Hosts.

2. Click the [+] icon at the top of the list.

3. In the Add Hosts dialog that opens, select a template from which to create a new host,

4. Specify the number of hosts to create. If the number you specify exceeds the Maximum hosts value set in the host pool properties (taking into account the number of VMs that already exist), you'll see a warning message and will need use a lower number or change the maximum host number on the Provisioning tab of the host pools properties.

5. Click OK to close the dialog.

6. After you click Apply in the RAS Console, the new hosts will appear in the list on the Desktop tab with the Status column saying "Cloning". Once the cloning is complete, the new hosts become available to users.

The Parallels RAS Test Template Wizard is used to test the health of a template. The wizard allows you to see that all post-prep activities for a template complete correctly. This includes checking DHCP settings, DNS registration, correct VLAN, joining the AD domain, correct target OU, etc.

To open the wizard, right-click a template in the Parallels RAS console and choose Test. The test procedure consists of the following steps:

1. The template is switched temporarily to the "Test" mode designed specifically for this purpose. Please note that while the template is in this mode, all other operations are blocked until the test is finished and the template exits the test mode.

2. A host is cloned from it to be used for testing. The VM is kept on the server for the duration of the test and will be deleted afterwards.

3. A series of tests is then run on the host to test the template from which it was created.

4. Once the test is complete, a report is displayed on the screen showing the test results.

When the wizard starts:

1. The Welcome page opens. Read the info that it contains and click Next when ready.

2. The next page displays the list of individual tests that will be performed, including:

   • Check host Agent: This test tries to communicate with the RAS Guest Agent installed in the VM. If the agent responds, it means that the VM has been created and started successfully.

   • Check domain membership: Checks that the computer has joined the AD domain.

   • Check target OU: Checks that the RDP connection to the computer is possible with domain credentials.

   • Launch Parallels Client: This test launches Parallels Client and establishes a connection with the host.

3. While the test is running, the progress indicator is displayed on the screen. If needed, you can cancel the test at any time by clicking the Cancel button.

4. Once all tests are completed, you will see a page displaying the test results:

   • Success: If all tests complete successfully, the temporary host will be marked for deletion and the template will be switched back to the normal operation mode.

   • Failure: If one or more tests fail, you will see the corresponding info and will be able to download the log file by clicking the Download log file link. You will also have an option to switch the template to maintenance mode, which will prevent creating hosts from it until it is fixed.

5. Click Finish to close the wizard.

Once the agent is installed, and the Status field on the Check Agent wizard page confirms this, click Next. The VM will now be powered off (wait for the power off operation to finish). Once the VM is powered off, the template configuration Step begins.

The subsequent wizard pages are described in the sections that follow this one. Please note that many of the wizard pages inherit the information from Site default settings, but you can override it if needed. To specify your own settings, clear the Inherit default settings option. To see and edit default settings, click the Edit Defaults link. For more information, see .

This section describes the host naming pattern that you specify on the page of the template creation wizard.

Each time a new host is created, a name for it is generated automatically based on the pattern that you specify in the Host name field. The complete name format is as follows:

<prefix>%ID:N:S%<ending>

where:

• <prefix> is an alphanumeric string that must begin with a letter (not a digit).

• %ID:N:S% is a numeric pattern used to automatically generate a unique host ID. See the Numeric pattern subsection below.

• <ending> is a free-form alphanumeric string.

Numeric pattern

The numeric pattern in the VM name has the following format:

%ID:N:S%

The elements in the pattern above are:

• ID — Must be included as is.

• N — The number of digits to use, including leading zeros. Use "0" if you don't want to insert leading zeros.

• S — The starting number. This element is optional. If you don't include it, the number will start with 1.

Examples:

• %ID:3% — This pattern will generate 3 digit numbers with leading zeros, such as "001", "002", "003"...."998", "999".

• %ID:3:200% — This example will generate 3 digit numbers starting from 200, such as "200", "201", "202"..."998", "999".

• VDI-R1-%ID:3:100% — This is a complete name with an alphanumeric prefix and a numeric pattern. The resulting names will look like the following: "VDI-R1-100", "VDI-R1-101", etc.

When crating a name pattern, follow the rules listed below. If any of these rules are not observed, you will see an error message and will have to correct it:

• The name must start with a letter. A digit is not allowed as the first character.

• The alphanumeric part of a name can contain letters, digits, and a hyphen. No other characters are allowed.

• The total length of the name must not exceed 15 characters.

• The name can include just one numerical pattern (%ID:N:S%), which must be placed at the end or in the middle of the name.

The pattern that you specify is also validated against the value of the Maximum hosts field. If the pattern doesn't cover the maximum number of hosts, you will get an error and will have to correct it.

Reusing VM numbers in a name

When you delete a host, the number that was assigned to it becomes unused. The next host that is created will be given this number, so there are no gaps in numbering.

To verify that a template is functioning as intended, you can examine its status in the main template list in the RAS Console (the Status column). When a template is functioning properly, the Status column displays "Ready", which means that hosts can be created from it as needed. When a template is being created or when it's in maintenance, or when it's being removed, the status will change accordingly.

Note that one of the other columns in the table is Agent status, which is the status of the RAS Guest Agent installed in a template. Compared to servers in the RAS infrastructure (Connection Broker, Gateways, RD Session Hosts, etc), the agent status is not as important in a template as the template status. This is because a template is not a regular virtual machine and is not normally running, so checking the agent status in a stopped VM doesn't make much sense. This is why the agent status for a template in the RAS Console is usually Not Available, which is perfectly normal. The only situation when the Agent status displays a meaningful value is when the template is in Maintenance and running like a regular VM, in which case.the agent is also running an its status can be verified.

The tables below describe what the Status and Agent status columns will show for various template states or transitions.

Template creation

Template in production

Template in maintenance

Template removal

If you need to change the configuration of an existing template, select it in the Templates list and click Tasks > Properties. This opens the Template Properties dialog, which consists of tabs containing the same properties as the wizard pages described in Step 2: Configure the template.

For the description of the Multi-Provider Template Distribution feature, please see the Multi-provider template distribution section.

Adding or removing Providers from the distribution list

You can add or remove a Provider to/from a distribution list at any time using the template Properties dialog. To open the dialog, right-click a template on the VDI > Templates tab and choose Properties.

Template distribution status

After you complete the Create Parallels Template Wizard and create a template, or when you add/remove a Provider to/from an existing template, you can monitor the template distribution status on the Templates tab. The status is displayed in the Distribution column and may have the following values:

• Distributing — the distribution is in progress (the template is being distributed to target hosts).

• OK — the template has been successfully distributed to all specified hosts.

• Removing / Adding provider — A Provider is being added or removed.

• Failed to distribute — indicates that an error has occurred during the distribution operation.

Distribution details

The Tasks > Distribution details menu on the Templates tab opens a dialog where you can view the current distribution State and Progress indicators for the Providers that use this template.

The Progress column displays the same values as the Distribution column in the main template list (see above).

The State column may display one of the following:

• Ready — The Provider is ready.

• Not available — The Provider is not responding.

• Needs update — The template distribution operation may need to be performed again. You can click the Retry button to retry the template distribution operation for this host.

Leaving maintenance mode

When a template leaves the maintenance mode, a prompt is usually displayed saying that "All hosts must be recreated because the template has been modified. Do you want to recreate them now?". If the administrator clicks Yes and the template uses multi-provider distribution, Parallels RAS verifies the status of each provider. If a Provider is not responding, a message is displayed, asking the administrator to check the provider status. You can bring the provider back online and try recreating the hosts again. If this cannot be done at this time, you can recreate the hosts later.

Hosts and other desktops are managed on the VDI > Desktops tab, where you can perform all of the standard desktop management operations from the Tasks menu. The operations include Recreate, Delete, Upgrade all Agents, Assign, Unassign, Show sessions, Start, Stop, Restart, Suspend, Reset, and others. The Restart operation (graceful) has a 10 min timeout. If not completed during this time, the Reset operation (forced) will be used.

By default, the Desktops tab displays all of the desktop available in the Farm (you may need to scroll the list to see all available desktops). To see just the hosts that belong to a specific template, select a template in the Templates tab and click Tasks > Show hosts. This will switch you to the Desktops tab where the list will be automatically filtered to include only the VMs that belong to the selected template.

For more information, see Managing hosts.

A host is called persistent when it is assigned to a particular user or device. To make a host persistent, do the following:

1. Begin publishing a desktop or a resource from a host.

2. When specifying Virtual Guest Settings options, select Enable static assignment to host.

3. Complete the publishing wizard.

4. As a result, the VM will be assigned to the first user or device who uses a desktop or a resource. For information on how to switch between user and device assignments see section Site defaults (VDI), subsection "General".

You can also manually assign a host to a user or device. To do so:

1. Navigate to Farm > <Site> > VDI > Desktops.

2. Select a host and click Tasks > Assign. You can assign a host to a user or device. For information on how to switch between user and device assignments see section Site defaults (VDI), subsection "General".

3. To assign to a user, specify the user registered in Active Directory.

4. To assign to a device, select one of the following options:

   • Add from Active Director allows you to add a device joined to a domain in Active Directory.

   • Add from known device allows you to add a device known to RAS Device Manager.

   • Add custom entry allows you to input the device name manually

5. As a result, the host will be assigned to the selected user.

To view persistent hosts, navigate to Farm > <Site> > VDI > Desktops. A persistent host is identified by the Persistent value in the Assignment column.

To remove persistence from a host, do one of the following:

• Select a host on the Desktops tab and then click Tasks > Unassign.

• Navigate to Farm > <Site> > VDI > Desktops and click Tasks > Site defaults. In the dialog that opens, use the Auto remove persistence if host was not used for option to select the time period after which persistence should be automatically removed. You can also type any desired time period, such as "1 week 3 days".

There are two basic types of hosts when using Parallels RAS VDI: template-based and non-template based. This topic describes management tasks for both host types, indicating whether a task applies to a particular host type.

Viewing host list

To view the list of non-template based hosts, select Farm > <Site> > VDI > Desktops. If you have a filter applied to the list, remove it by click the magnifying glass icon. Without the filter, the list shows all desktops available in this RAS Farm, including hosts (both template-based and non-template based), hosts from a pool (RAS or native), and pool-based Remote PCs. Therefore, the Desktops tab is a location where you can view all of your desktops in one place. Here you can perform all of the standard desktop management tasks accessible from the Tasks menu, including Recreate, Delete, Assign, Unassign, Start, Stop, Restart, Suspend, Reset, Show sessions, and others.The Restart operation (graceful) has a 10 min timeout. If not completed during this time, the Reset operation (forced) will be used.

To view the list of hosts created from a template, select Farm > <Site> > VDI > Templates. Select a template and click Tasks > Show hosts. You will be switched to the Desktops tab where the list of desktops will be filtered to include only those that belong to the template. As was mentioned above, you can perform all of the standard desktop management operations on this tab, including power operations, which are described in detail later in this section.

For the list to include only the hosts from a particular pool, select a pool in the Pools tab and click Tasks > Show hosts in Pool.

The filter in the Desktop tab can also be applied manually by clicking the magnifying glass icon and entering the filter criteria in the fields that appear at the top of the list.

Site defaults

Hosts created from a template inherit the template settings. To view the settings, note on which template a host is based and then view properties of that template, specifically the Settings and Security tabs. For more information, see Site defaults. Note that you a template can inherit Site default settings or you can specify your own custom settings for it.

Non-template based hosts have their own settings, some of which (specifically Settings and Security) are inherited from Site defaults. To see settings for a non template-based VM, navigate to Farm > <Site> > VDI > Desktops. A host that doesn't belong to a template is identified by an empty value in the Template column. Right-click a template and choose Properties (note that template-based hosts do not have this menu option).

Checking the RAS Guest Agent status

A host must have the RAS Guest Agent installed and the agent must match the Parallels RAS version. The agent is installed by default when a host is created from a template. If a host was created using the native hypervisor tools, it may not have the agent installed in it. In such a case, the host will be able to serve only the remote desktop. To enable it to server applications or documents, you'll need to install the agent yourself.

To check if the RAS Guest Agent is installed in a host and is up to date:

1. Select a host in the list and then click Tasks > Troubleshooting > Check agent.

2. The Guest Agent Information dialog opens displaying the information about the RAS Guest Agent.

3. If the agent is not installed, click the Install button and follow the instructions. The agent will be push installed in Windows running inside the host.

Deleting a host

To delete a template-based host, select it and then click the Tasks > Delete.

Managing hosts that failed preparation

If a template-based host encounters a problem during the preparation stage, it remains on the server but cannot be used. You can identify such VMs by the "Failed to create" value in the Status column. Unless a VM like this is repaired, it will be automatically removed after the time period specified in the Site defaults. To see Site defaults:

1. Select Farm > <Site> > VDI > Desktop and then click Tasks > Site defaults.

2. In the dialog that opens, on the General tab, view or modify (if needed) the Auto remove hosts which failed preparation after option. You can set any of the available time periods by selecting it from the drop-down list or you can type a desired value, such as "8 days" or "12 hours".

Recreating a host

If something happens to a template-based host and it becomes unusable, you don't have to delete it and create a new one. Instead, you can recreate it keeping its name, MAC address, and other properties. This way none of the other Site settings, which may rely on a broken host, will be affected. Another reason for recreating a host is to apply changes made to the template (when you exit from maintenance without executing the Recreate command).

Please note that recreated VMs can keep the the following properties:

• MAC address is kept on ESXi, vCenter, Hyper-v, Hyper-v Failover Cluster, Nutanix AHV (AOS), and SC//HyperCore.

• BIOS UUID is kept on ESXi and vCenter.

• DRS groups are kept on vCenter.

To recreate one or more hosts:

1. In the Parallels RAS Console, navigate to Farm > <Site> > VDI > Templates.

2. To recreate all deployed hosts, click the Tasks drop-down list and choose Recreate all hosts.

3. To recreate a specific host (or multiple hosts), click Tasks > Show hosts. This will open the Desktops tab, which will list hosts. Select one or more hosts and then click the Tasks > Recreate.

When you recreate a host:

• The procedure deletes a VM and creates a new one from the same template.

• The new host retains the same computer name as the one it replaces.

• If a host is running, all unsaved data in its memory will be lost. For this reason, an important data should be saved to an external storage.

A template can be put into to a special mode called "maintenance", which is primarily used to update or install software in the guest operating system. While in this mode, the template becomes unavailable for all normal tasks, including creating new hosts, and it becomes possible to start it as a regular virtual machine. Once the virtual machine is running, you can install or update software in the guest OS or perform administrative tasks in the operating system.

Depending on whether a template is configured for full or linked clones, the maintenance mode is used slightly differently, as described below.

Full clones

If your template is configured to create full clones, do the following:

1. Select a template and click Tasks > Maintenance and select the template version to be put in maintenance. The template becomes disabled (grayed out) and all operations on it are suspended. The status of template in the Status column changes to Entering maintenance and when completed, changes to Maintenance.

2. Using native tools of the hypervisor, start the template as a normal virtual machine.

3. Install Windows updates or software as necessary.

4. When done, shut down the virtual machine.

5. Back in the RAS Console, select the template and click Tasks > Maintenance again to exit the maintenance mode. You will see a dialog where you can choose whether to create a new version or discard the changes. Select Create a new version.

1. Create a new template version as described in section Using template versions, subsection "Creating a new version". When you update a full clone template, the changes will only affect future clones. For existing clones to have these updates, they must be recreated. You can choose to recreate existing hosts now or you can postpone it. Please note that recreating a full clone is a time consuming process. Also, a new app may be installed in a full clone VM or a user profile may be changed while the recreation is in progress, all of which will be lost. To minimize impact on users, it makes sense to schedule a maintenance window during which the clones can be recreated.

Linked clones

Since linked clones share the virtual hard disk with a snapshot of a template, you need to take additional steps compared to full clones.

First, you need to notify host users to save their data and log off. This is necessary for existing hosts to include the updates that you will install in the template. Once all users are logged off, do the following:

1. Select a template and click Tasks > Maintenance and select the template version to be put in maintenance. The template becomes disabled (grayed out) and all operations on it are suspended. The status of the operation is displayed at the bottom of the window.

2. Using native tools of the hypervisor, start the template as a normal virtual machine.

3. Install Windows updates or software as necessary.

4. When done, shut down the virtual machine.

5. Back in the RAS Console, select the template and click Tasks > Maintenance again to exit the maintenance mode. You will see a dialog where you can choose whether to create a new version or discard the changes. Select Create a new version.

1. Create a new template version as described in section Using template versions, subsection "Creating a new version". Please note that if you create a new version without recreating linked clones, you will have to recreate them manually or by using scheduler.

Updating RAS Guest Agent inside a template

A template must have the latest version of RAS Guest Agent installed in it. The agent is installed when you create a template. When a new version of RAS Guest Agent becomes available, it should be updated. To update the agent, the maintenance mode must be used as described above. To simplify agent updates, Parallels RAS monitors all installed agents and notifies the administrator when an update is available.

When the RAS Console starts, all installed agents are checked and a message is displayed if one or more agents need to be updated. This applies to servers in the RAS infrastructure and the templates. The message will ask if you want to update all agents. If you click Yes, you are presented with a dialog listing all servers and templates on which an agent needs to be updated. You can select or un-select a server/template to include it in the bulk update procedure or exclude it. Once you've made your selection, click OK to start the update. Follow the onscreen instructions and update the agents.

When you update the Agent in a linked clone template, you'll be asked if you want to recreate all hosts that were created from this template. You can click Yes and they will be automatically recreated to match the template.

When you update the Agent in a full clone template, full clone hosts are not automatically recreated. You will be asked if you want to recreate them. If you decide to do so, please note that full clone VMs are complete machines, so recreating them is a time-consuming process. Alternatively, you can update the agent in these VMs by push-installing it from the RAS Console. This can be done by clicking Tasks > Upgrade all Agents while on the VDI > Desktops tab.

To manually check the RAS Guest Agent status in a template, click Tasks > Check agent. If the agent is up to date, a message box is displayed confirming this. If a newer version of RAS Guest Agent is available, you'll see a dialog asking you to update it. Please note that the difference in updating full and linked clone templates (as described above) applies to this scenario as well.

Maintaining RD Session Hosts based on a template

If you need to do a scheduled maintenance of RD Session Hosts that were created from a template, please follow these steps:

1. Create a schedule that fits your maintenance window to drain a desired RD Session Host group.

2. During maintenance (or right before it) switch the template into maintenance mode. Then apply the necessary changes.

3. The schedule disables groups provisioned by the template (while the maintenance window lasts) which leads to removing (unassigning) all hosts from them.

4. Release the template from maintenance and click Yes when asked whether to recreate all clones.

5. Enable groups which were disabled in step 3 (above). At this point, the groups will begin receiving hosts to comply with Keep Available Buffer setting

6. From this point forward, groups are provisioned with VMs on demand.

Please see Session Management.

High availability for VDI means that a Provider must never lose a connection with a Provider Agent. If the connection is lost, the hosts will become unavailable for user connections. High availability for VDI is accomplished by installing at least three RAS Connection Brokers. This way, if one of the Connection Brokers goes offline (and with it the built-in Provider Agent), the Provider will be automatically assigned to the Provider Agent running in the next available Connection Broker.

To configure high availability for VDI, use the information and instructions below.

At least three Connection Brokers are required

Make sure you have at least three RAS Connection Brokers installed and running. When RAS Connection Brokers from your site are online, high availability is enabled automatically. You may also have additional Connection Brokers in standby mode, but you must have at least three agents in the active state for the high availability functionality to work. All Connection Brokers must be able to communicate with each other.

An odd number of agents is recommended

To properly control a possible split-brain situation, strictly more than half of all available Connection Brokers should be able to communicate with each other at any given time. Consider the following examples:

• Let's say there are three Connection Brokers in a Site. All of them can communicate with each other. If one of the agents suddenly loses a connection with the other two, the two agents will know that they are in the majority and will take over the Provider hosts that are currently managed by the first agent.

• Let's now say that there are four Connection Brokers. If one of them loses a connection to the remaining three, the same scenario will occur as in the example above. But if two agents simultaneously lose a connection to the other two, none of the two groups will be in the majority and therefore none will be able to make a decision who should take over the Provider hosts. In a situation like this, steps must be taken to prevent a split-brain scenario, which will happen if the agents continue to operate independently from each other. As a solution to this problem, all agents will simply abandon all Providers at the same time, so no data loss or any other problem can possibly happen.

For the reasons explained above, you should always install an odd number of Connection Brokers. This way, one of the groups of agents will always be in the majority and will continue to handle all Providers. Please note that the general recommendation (regardless or the high availability functionality described here) is to have three RAS Connection Brokers running in a Site. For details, see Secondary Connection Brokers.

Please also note that Connection Brokers in standby mode don't participate in the high availability operations. These agents stay inactive until one of the active Connection Broker goes completely offline. When that happens, an agent in standby mode is activated and takes place of the lost agent. From this point forward, it is considered a part of the high availability setup. When the lost agent is brought back online, everything goes back to what it was before.

Configuring a Provider for high availability

Parallels RAS can maintain high availability in the following ways:

• Parallels RAS automatically selects a Connection Broker for the Provider. If this Connection Broker goes down, Parallels RAS moves the workload to one of the remaining Agents.

• You select a Connection Broker for the Provider manually. If this Connection Broker goes down, Parallels RAS moves the workload to one of the remaining Agents.

Use one of the following to configure a Provider for high availability:

• For an existing Provider, open the Properties dialog, select the Agent Settings tab and in the Preferred Connection Broker field, select Automatic or manually select your preferred Connection Broker.

• When adding a new Provider, on the second wizard page where you specify the host type and address, click the Advanced Settings link and in the Preferred Connection Broker drop-down list select Automatically or manually select your preferred Connection Broker. Note that the Automatic option is selected by default when there are three or more Connection Brokers available.

You can perform standard computer management tasks on a server right from the RAS Console. The tasks include Remote Desktop Connection, PowerShell, Computer Management, Service Management, Event Viewer, IPconfig, Reboot, and others. To access the Tools menu, select a server, click Tasks > Tools and choose a desired tool. For the complete description, please see Computer management tools.

In addition to the Provider editor described earlier in this chapter, you can also see summary about the available Providers. To do so:

1. In the RAS Console, select the Farm category and then select the Site node in the middle pane.

2. The available servers are displayed in the VDI section in the right pane.

3. To go to the Provider editor, right-click a server and choose Show in the Editor.

For additional info, see Sites in the RAS Console.

The Scheduler tab allows you to create scheduler tasks that will be performed on individual hosts or host pools at a specified time.

Disabling hosts and hosts in pools

To disable a host or a host in a pool:

1. Click Tasks > Add > Disable host or Disable host pool.

2. On the General tab, select the Enable Schedule option.

3. Specify a name for this schedule and an optional description.

4. Select a host or a pool in the Available list and click Add. The host or pool will appear in the Target list.

5. Select the Trigger tab and specify start date and time, duration, and recurrence settings for this event. To make this a one-time event, select Never in the Recur drop-down list.

6. Select the Options tab. It contains the following options:

   • Message list: Configure a message that will be sent to users before the host goes offline. Click Tasks > Add and specify the message title, body, and the time period when it should be sent.

   • On disable: Specify what should happen to current sessions when a scheduled task triggers. Select the desired option from the On disable drop-down list.

   • Enforce schedule for currently inactive hosts: This option is only enabled when you have an active message in the list. If the option is enabled, hosts that are currently offline are also monitored, and if such a server comes back online during the scheduled task execution, the task is applied to it too.

7. Click OK to save the schedule.

Rebooting hosts and hosts in pools

To reboot a host or a host in a pool:

1. Click Tasks > Add > Reboot host or Reboot host pool.

2. On the General tab, select the Enable Schedule option.

3. Specify a name for this schedule and an optional description.

4. Select a host or a pool in the Available list and click Add. The host or pool will appear in the Target list.

5. Select the Trigger tab and specify start date, time, and recurrence settings for this event. To make this a one-time event, select Never in the Recur drop-down list. In addition, specify the following options for the "Reboot host pool" task:

   • Complete in: Specify the time to complete the task.

6. Select the Options tab. It contains the following options:

   • Message list: Configure a message that will be sent to users before the host is rebooted. Click Tasks > Add and specify the message title, body, and the time period when it should be sent.

   • Enable Drain Mode and Force server reboot after: The two options work together. If you select the Enable Drain Mode option, then when the task triggers, new connections to a host are refused but active connections will continue to run and can be reconnected. The server will be rebooted when all active users close their sessions or when Force server reboot after time is reached, whichever comes first. For active users not to lose their work, create a message that will advise them to save their work and log off.

   • Enforce schedule for currently inactive hosts: This option is only enabled when you have an active message in the list. If the option is enabled, hosts that are currently offline are also monitored, and if such a server comes back online during the scheduled task execution, the task is applied to it too.

7. Click OK to save the schedule.

Starting up hosts and hosts in pools

To start up a host or a host in a pool:

1. Click Tasks > Add > Startup host or Startup host pool.

2. On the General tab, select the Enable Schedule option.

3. Specify a name for this schedule and an optional description.

4. Select a host or a pool in the Available list and click Add. The host or pool will appear in the Target list.

5. Select the Trigger tab and specify start date, time, and recurrence settings for this event. To make this a one-time event, select Never in the Recur drop-down list.

6. ("Startup host pool" task only) Select the Options tab. It contains the following options:

   • Power on all members: Select this option to start up all hosts assigned to specific users.

   • Percentage of members: Select this option to specify the percentage of hosts that must be started up in each pool.

   • Specific number of members to be started: Select this option to specify the number of hosts that must be started up in each pool.

7. Click OK to save the schedule.

Shutting down hosts and hosts in pools

To shut down a host or a host in a pool:

1. Click Tasks > Add > Shutdown host or Shutdown host pool.

2. On the General tab, select the Enable Schedule option.

3. Specify a name for this schedule and an optional description.

4. Select a host or a pool in the Available list and click Add. The host or pool will appear in the Target list.

5. Select the Trigger tab and specify start date, time, and recurrence settings for this event. To make this a one-time event, select Never in the Recur drop-down list.

6. Select the Options tab. It contains the following options:

   • Message list: Configure a message that will be sent to users before the host is rebooted. Click Tasks > Add and specify the message title, body, and the time period when it should be sent.

   • Enable Drain Mode and Force server shutdown after: The two options work together. If you select the Enable Drain Mode option, then when the task triggers, new connections to a host are refused but active connections will continue to run. The server will be shut down when all active users close their sessions or when Force server reboot after time is reached, whichever comes first. For active users not to lose their work, create a message that will advise them to save their work and log off.

   • Enforce schedule for currently inactive hosts: This option is only enabled when you have an active message in the list. If the option is enabled, hosts that are currently offline are also monitored, and if such a server comes back online during the scheduled task execution, the task is applied to it too.

Recreating hosts and host pools

To recreate a specific host or all hosts in a host pool:

1. Click Tasks > Add > Recreate host from template or Recreate host pool from template.

2. On the General tab, select the Enable Schedule option.

3. Specify a name for this schedule and an optional description.

4. Select a host or a pool in the Available list and click Add. The host or pool will appear in the Target list.

5. Select the Trigger tab and specify start date, time, and recurrence settings for this event. To make this a one-time event, select Never in the Recur drop-down list.

6. Select the Options tab. It contains the following options:

   • Message list: Configure a message that will be sent to users before the host is rebooted. Click Tasks > Add and specify the message title, body, and the time period when it should be sent.

   • Enable Drain Mode, Force host recreation after, and Force host pool recreation after: The options work together. If you select the Enable Drain Mode option, then when the task triggers, new connections to a host are refused but active connections will continue to run. The server will be recreated when all active users close their sessions or when the time specified in Force host recreation after or Force host pool recreation after is reached, whichever comes first. For active users not to lose their work, create a message that will advise them to save their work and log off.

   • Enforce schedule for currently inactive hosts: This option is only enabled when you have an active message in the list. If the option is enabled, hosts that are currently offline are also monitored, and if such a server comes back online during the scheduled task execution, the task is applied to it too.

Sending multiple messages to users

For Disable Host and Disable Host Pool tasks, you can only send a message before the scheduled task is triggered. Hence, when creating a message, you can only select the "before" option when specifying when the message should be sent. You can create more than one message if needed and send them at different time intervals, so the users are notified more than once before the task executes.

For Reboot Host and Reboot Host pool tasks, you can send a message before or after the scheduled task is triggered. The "after" option is available for these tasks because you have the ability to enable the drain mode, which will keep the active sessions running for some time. During this time, you can send multiple messages to active users reminding them that they should finish their work and close their sessions. To use the "after" option, the Enable Drain Mode option must be selected. Please also note that the "after" time interval and the Force server reboot after setting should be coordinated. For example, if the force reboot occurs before the "after" time elapses, active users will not have a chance to see the message.

Site defaults are settings that are defined on a Site level and can be used by templates and hosts (both template-based and non-template based). By default, templates (described later in this chapter) inherit Site default settings, but you can override them if needed when you configure a template. Non-template based hosts also use Site default settings by default and you can also override them if needed when you configure these VMs.

To view and modify Site defaults, do the following:

1. Navigate to Farm > <Site> > VDI.

2. Select the Desktops tab in the right pane.

3. Click Tasks > Site defaults. This opens the Site Default Properties dialog, which is described below.

Note that any modifications you make to Site defaults are immediately applied to all hosts in the current Site that use them.

General

The General tab contains the following properties:

• Session readiness timeout: The maximum amount of time it should require to establish a session. If the specified timeout is reached, and the session is still not ready, the user will see an error message and will have to try to log in again.

• Protocol: Specifies a protocol that Parallels RAS uses to communicate with a host.

• Auto remove hosts which failed preparation after: If a host encounters a problem during the preparation stage (for any reason), it remains on the server but cannot be used. You can identify such VMs by the "Failed to create" value in the Status column (Farm > <Site> > VDI > Desktops). Unless a VM like this is repaired, it will be automatically removed after the time period specified in this field. You can set any of the available time periods by selecting it from the drop-down list or you can type a desired value, such as "8 days" or "12 hours".

• Desktop assignment type: Specifies whether the persistent hosts are assigned by the UPN (the User option) or device hostname (the Device option). Each host will be automatically assigned to the first user or device who uses a resource published from it with persistent assignment enabled. You can also assign hosts manually. For more information, see Persistent Hosts.

• Auto remove persistence if guest was not used for: The time period after which persistence should be automatically removed. You can also type any desired time period, such as "1 week 3 days".

User profile

Configure this tab as descried in User profile.

Application Packages

Configure this tab as descried in Using MSIX application packages.

Optimization

Configure this tab as descried in Optimization.

Actions

Actions: The two drop-down lists here specify an action to perform on session disconnect or logoff.

Security

On the Security tab, you can specify whether to automatically grant users Remote Desktop connection permissions on hosts. Here's how it works. Instead of manually adding each user to the Remote Desktop Users (or Administrators) group, you can enable this option to do it automatically. When a user logs on, he/she will be automatically added to the specified group and will therefore have the Remote Desktop connection (or full Administrator) permissions on the server. When the user logs off, they will be removed from the group (i.e. the group membership will only exist for the duration of the session).

The more important benefits of this feature are as follows:

• You don't have to permanently add your users to the Remote Desktop Users groups. This way, a user will never be able to establish a Remote Desktop session with a server outside of Parallels Client.

• By automatically adding a user to the Administrators group, you can give them rights to install applications and perform other administrative tasks. Once again, the user will only be able to do it from Parallels Client but never by connecting to the server using standard Remote Desktop tools.

Settings

The Settings tab contains the following:

• Disconnect active session after: The amount of time a session remains logged in after the user closes a published application. The default timeout is 25 seconds. Note that this only works for applications, but not published desktops (when a user closes a desktop, the session is logged off). This timeout is used to avoid unnecessary logins when a user closes one application and then opens another.

• Preferred Connection Broker: Select a preferred Connection Broker to which this Provider should be assigned. This can be helpful when Site components are installed in multiple physical locations communicating through WAN. You can decrease network traffic by specifying a more appropriate Connection Broker.

• Allow URL/Mail redirection: When a user tries to open a URL or an HTML Mailto link in a remote application, the link can be redirected to the client computer and open in a local default application (a web browser or email client) instead of an application on the remote host. This option allows you to enable or disable the redirection. Click the Configure button to choose from the following options:

  1. Replace registered application — this option uses an alternative method of redirecting a link. It replaces the default web browser and mail client with "dummy" apps on the remote server side. By doing so, it can intercept an attempt to open a link and redirect it to the client computer.

  2. Support Windows Shell URL namespace objects — the Shell URL namespace objects support means that Parallels RAS can intercept actions in published applications that use Shell namespace API to open links, which is a standard behavior in most applications. The ability to disable the support for Shell URL namespace objects is for compatibility with older versions of Parallels RAS.

• Enable drag and drop: Allows you to set how the drag and drop functionality works in Parallels Clients. Click Configure and choose from "Disabled" (no drag and drop functionality), "Sever to client only" (drag and drop to a local application only), "Client to server only" (drag and drop to a remote application only), "Bidirectional" (drag and drop in both directions).

• Manage RDP transport protocol: Selects the transport protocol that will be used for connections between Parallels Client and a server. To do this, select this option and click the Configure button.

• Allow file transfer command (Web and Chrome clients): Enables file transfer in a remote session. To enable file transfer, select this option and click the Configure button. For more information, see Configuring remote file transfer.

• Enable drive redirection cache: Improves user experience by making file browsing and navigation on redirected drives much faster. For details, see Drive redirection cache.

RDP printer

The RDP Printer tab allows you to configure the renaming format of redirected printers. The format may vary depending on which version and language of the server you are using. Select the RDP Printer Name Format option specifically for the configured server:

• Printername (from Computername) in Session no.

• Session no. (computername from) Printername

• Printername (redirected Session no)

The other RDP Printing option available is Remove session number from printer name, which will do what it says.

A persistent Remote PC is a PC assigned to a particular user. Once a PC is assigned, no other user can connect to it.

There are two ways to make a Remote PC persistent:

• When you publish a resource (application, desktop, etc.) from a pool-based Remote PC using the publishing wizard, you can select the Persistent option in the Virtual Guest Settings section. This way, a Remote PC in a pool will be assigned to the first user that opens the published resource. For more info, see Publishing from a pool-based Remote PC.

• You can also assign a Remote PC to a user manually. To do so, navigate to Farm > <Site> > VDI, select the Desktops tab, then select a Remote PC in the list and click Tasks > Assign. In the dialog that opens, specify the target user.

To remove persistence from a Remote PC, select it in the Desktops tab and click Tasks > Unassign.

RAS Guest Agent installed on a host discovers RAS Provider Agent by sending broadcast packets over UDP port 30006 to all devices in its subnet.

The discovery packet won't be received if the Agents are on different broadcast domains. In this case, you need to specify the Provider Agent's IP through the registry.

To specify the Provider Agent's IP:

1. On the host, open the Windows registry.

2. Add a new string value called 2XVDIAgent at:

   • 32-bit systems:HLKM > Software > Parallels>GuestAgent

   • 64-bit systems: HLKM > Software > WOW6432Node > Parallels>GuestAgent

The value should be specified as follows:

• If a dedicated RAS Provider Agent is used, the value must be set to the FQDN or IP address of the server where the Agent is installed.

• If the built-in RAS Provider Agent is used with manual agent selection, the value must be set to the FQDN or IP address of the RAS Publishing Agent.

• If the built-in RAS VDI Agent is used and the Agent is selected automatically (high availability), the string must contain FQDNs or IP addresses of all RAS Publishing Agents separated by a semicolon (i.e., <PA1 address>;<PA2 address>;<PA3 address>).

1. Restart the RAS Guest Agent service on the host to apply the changes.

Use the Preparation page to select and configure an image preparation tool.

First, select whether you want to use RASprep or Sysprep. The advantages of using RASprep and the differences between the two tools are described below.

RASprep is the Parallels RAS tool for preparing Windows in a VM after cloning it from a base image. RASprep performs the following tasks during the initial startup of each new VM:

• Creates a new computer account in Active Directory for each host.

• Gives the host a new name.

• Joins the host to the Active Directory domain.

Compared to Sysprep, RASprep works much faster because it modifies a lower number of configurable parameters and requires less reboots.

The following table lists the main differences between RASprep and Sysprep:

After selecting the preparation tool, specify the following options:

• Computer name: A name pattern that should be used to assign a computer name. For example, Windows10-RAS-%ID%.

• Owner name: Owner name (optional).

• Organization: Organization name (optional).

• Administrative password: Local Windows administrator password.

• Join domain: A domain name for the VM to join.

• Administrator: Domain account.

• Password: Domain account password.

• Target OU: Full DN of an organizational unit. Click the [...] button to browse Active Directory and select an OU.

Remote PC pools is a Parallels RAS feature that allows you to create pools of standalone (preferably domain-joined) PCs and optionally assign them to specific users. The Remote PC pools functionality is integrated into RAS VDI to take advantage of the infrastructure that already handles host pools.

Remote PC pools vs. Remote PCs

Remote PCs are standalone machines (physical or virtual) that can be used to host published resources in Parallels RAS. Remote PCs are managed in the Parallels RAS Console in the Farm > <Site> > Remote PCs section. The chapter describes this functionality in detail. Remote PC pools described here are handled separately and differently from standalone Remote PCs. They are managed in the Farm > <Site> > VDI section of the RAS Console.

You can add Remote PCs to a Provider in two ways:

• In the wizard while adding the provider as described in

• After creating the provider, as described in this section.

To add Remote PCs to a Provider after it has been created:

1. Right-click a Provider that you created and choose Properties.

2. Select the Remote PCs tab.

3. Add Remote PC as described in steps 9 and 10 in .

To set up a Remote PC pool in the RAS Console, you first need to add a Provider of type Remote PC. This is a special type that exists specifically for the purpose of creating and managing Remote PC pools. It is not a real Provider, so it doesn't need a hypervisor installed. It simply uses the existing VDI functionality to create and manage computer pools. Note that when you add a Provider of this type, you can manage it like any other real Provider with some limitations, such as you cannot create templates and use some other strictly VDI (hypervisor)-specific functionality.

To add a Provider of type Remote PC:

1. Navigate to Farm > <Site> > Providers.

2. On the Providers tab, click Tasks > Add.

3. Select one of the following:

   • Remote PC dynamic: This approach assigns PCs using the information from Active Directory. All you have to do is specify an organizational unit (OU) containing computer accounts to be assigned to the host.

   • Remote PC static: Using this approach, Remote PCs are assigned to the Provider by entering their FQDN or IP address (one by one) or by importing a list from a CSV file.

4. In the wizard that opens, specify the following:

   • Name: Name of the provider.

   • Description: Description of the provider.

   • Address: FQDN or IP address of a server that will manage Remote PC pools. This must be a server with RAS Provider Agent installed. You can use the RAS Connection Broker server, since it has the RAS Provider Agent built in, but it can be any other server running a dedicated RAS Provider Agent.

   • Username: Account name in the UPN format (e.g. administrator@domain.local). This must be a domain user account with administrative rights on the server specified above. Using a local Windows account is also possible with some limitations and only when using the static PC assignment (see below). Using a domain account is recommended.

   • Password: Account password and an optional description.

5. Click the Manage Credentials button to specify the accounts that will be used to deploy RAS agents.

6. Click the Advanced Settings link to open the Advanced Provider Settings dialog. The dialog allows you choose the following options:

   • Use dedicated Provider Agent: Select this option if you will install (or have installed) the RAS Provider Agent yourself. Clear the option if you will use the .

   • Agent address: This option becomes enabled if you select the option above it. Specify the FQDN or IP address of the server where the RAS Provider Agent is (or will be) installed. This can be either a physical box or virtual machine.

   • Preferred Connection Broker: Select a RAS Connection Broker to be the preferred agent for this Provider. For more info, see .

7. Click Next.

8. The wizard will display the new Provider information and will indicate the RAS Provider Agent status. If everything is OK, click Next.

9. If you selected Remote PC dynamic in step 3, specify the following:

   • Target OU: Organizational unit (OU) containing computer accounts to be assigned to the host. You can click the [...] button to browse Active Directory. Note that a maximum of 1000 Remote PCs can be returned in a single AD/OU search result.

     Note: When using dynamic assignment, Remote PCs must be domain-joined. You cannot manage such PCs using a local Windows user account.

     When you use the dynamic assignment, you have an option to install RAS Guest Agent on every PC by adding a Group Policy to the organizational unit with a script to deploy RAS Guest Agent. The following is an example of such script:

     msiexec /i RASInstaller-<version & build>.msi ADDLOCAL=F_GuestAgent /qn+ /norestart

     Other agent installation options are described in .

   • Subnet mask: Subnet mask used for calculating the directed broadcast address from the IP adderesses of the Remote PCs. It is used for sending a directed broadcast of the Wake on LAN magic packet.

10. If you selected Remote PC static in step 3, do one of the following:

    • Click Tasks > Add and type FQDN or IP address of a PC you want to add. You can click the [...] button to search for it. Next, specify the subnet mask used for calculating the directed broadcast address from the IP address of the Remote PC. It is used for sending a directed broadcast of the Wake on LAN magic packet. After this, enter the MAC address of the computer you are adding. Note that all fields are mandatory.

    • Click Tasks > Import from CSV file and then select a CSV file containing the list of computers. The CSV file must contain two columns: (1) FQDN or IP address; (2) MAC address. Once again, both columns are mandatory and must contain valid values.

    Parallels RAS 18 (and newer) supports a maximum of 1000 Remote PCs per Provider.

• Click Finish.

Management of pool-based Remote PCs includes assigning a PC to a specific user, upgrading the RAS Guest Agent, viewing and modifying PC properties, performing some standard administrative tasks, and some others.

To manage Remote PCs in a pool:

1. In Farm > <Site> > VDI, select the Desktops tab.

2. Note that the list on this tab includes all managed desktops, including hosts and pool-based Remote PCs. You can order the list by the Pool column to see Remote PCs assigned to a particular pool.

3. Select a Remote PC, click the Tasks drop-down list and choose one of the options described below. Note that not all options available in the Tasks menu are applicable to Remote PCs. The list below describes only the options that you can use with pool-based Remote PCs.

The Tasks menu options that apply to Remote PCs are:

• Upgrade all Agents. Upgrade RAS Guest Agent in all Remote PCs (and hosts) in the list.

• Assign. Assign a Remote PC to a specific user (make a PC persistent). Click the menu option and specify a user.

• Unassign. Remove the user assignment (persistence) from a Remote PC.

• Show sessions. Switches the view to the Sessions tab and displays the session information.

• Tools. Allows to perform a set of standard operations, such as establishing a remote desktop connection, pinging, rebooting/shutting down a Remote PC, and others. For the description of power operations, please see Performing power operations below.

• Troubleshooting. Check and install/upgrade the RAS Guest Agent in a Remote PC.

• Reset properties. Resets Remote PC properties to their default values. See Properties below.

• Properties. Opens a dialog where you can view and modify Remote PC settings. The General tab allows you to temporarily disable the Remote PC in a pool (use the Do not use this host option). This is specifically useful when you need to perform maintenance tasks on a PC. You can also view and modify the Remote PC display name, computer name, and the port number on which it communicates with the Provider. For the description of Settings and Security tabs, see .

Performing power operations

For remote power operations to work, WMI must be enabled in Windows running in a VM and TCP ports 30004 and 30005 must be open. At the time of this writing, this is not automated in Parallels RAS but will be in future versions.

To perform power operations on a host (start, stop, restart, suspend, reset), open the VDI > Desktops tab, select a host, then click Tasks and choose an operation that you want to perform (for start and stop operations, you can click the corresponding icons at the top). The restart operation (graceful) has a 10 min timeout. If not completed during this time, the reset operation (forced) will be used.

Please note if you are using Nutanix AHV (AOS), the suspend operation is not available (the Suspend icon is disabled). The reason for this is Nutanix AHV (AOS) does not support the suspend operation on its virtual machines.

To configure logging and retrieve or clear existing log files, right-click a Provider, choose Troubleshooting > Logging in the context menu, and then click one of the following, depending on what you would like to do: Configure, Retrieve, or Clear. For the information on how to perform these tasks, see the section. Please also read the important information below.

Note that logging of Provider operations is performed on the RAS Provider Agent level. When you configure logging for a Provider, you are essentially configuring it for the RAS Provider Agent that services this Provider. This means that if you are using the built-in RAS Provider Agent, its logging configuration applies to all Providers that it services. Consider the following scenarios:

• When you retrieve log files for a specific Provider serviced by the built-in Provider Agent, the files will contain logs for all Providers serviced by the same agent.

• If you clear log files for a particular Provider, you should be careful because the logs will be cleared for all Providers if they are serviced by the same built-in Provider Agent. The RAS Console will prompt you if you try to delete such a shared log.

If a Provider has a dedicated Provider Agent, which services this host only, none of the above applies.

To be managed in a Remote PC pool, a Remote PC must have RAS Guest Agent installed. This can be done using one of the following options:

• When you add an individual Remote PC to a pool, you'll be asked to upgrade the agent. Follow the onscreen instructions and install or upgrade it.

• When you add all Remote PCs in a host to a pool at once, you can add them first and then use the Tasks > Upgrade all Agents menu option in the Desktops tab.

• When you assign Remote PCs to a Provider via Active Directory, you can have a Group Policy in the OU with a script to deploy the agent. See .

• To install or upgrade the agent on an individual Remote PC, select it in the Desktops tab and click Tasks > Troubleshooting > Check agent option. In the dialog that opens, click Install.

• Finally, you can install RAS Guest Agent manually by running the Parallels RAS installer on a Remote PC and selecting to install the RAS Guest Agent component.

Once you assigned PCs to a Provider, you can add them to a Remote PC pool as follows:

1. In Farm > <Site> > VDI, select the Pools tab.

2. Add a new pool by clicking Tasks > Add in the Pools pane.

3. Select the pool that you've created and then in the Members pane, click Tasks > Add and choose one of the following:

   • All Hosts in Host: Adds all Remote PCs assigned to the Provider. When you click this option, a dialog opens allowing you to select a Provider. Select the host and click OK.

   • Host: Adds an individual Remote PC. In the dialog that opens, select a desired Remote PC and click OK. Another dialog may open asking you to upgrade RAS Guest Agent on a Remote PC (the agent is required for a PC to be managed in a pool). Click OK to upgrade (or install) the agent. You can also upgrade the RAS Guest Agent on one or more PCs at another time as described in .

Once you add one or more Remote PCs to a pool, they will appear in the Pool management tab and in the Desktops tab.