Citrix LAB Single Deployment

Fork of Ryan Butler github, however this version just deploys a single server with all the roles (Director, Delivery Controller, Storefront, SQL, and VDA).

What to edit

• Make Edits/Change Variables to the following files
  • Root
    • Build.sh - Change directory to project location
    • Destroy.sh - Change directory to project location
  • Terraform
    • Labtf.vars
  • Ansible
    • Vars.yml
    • Roles
      • Site-hydrate
        • defaults\main.yml
      • Storefront-init
        • defaults\main.yml
        • files\cert.pfx (drop ssl cert file here)

Citrix Virtual Apps and Desktop vCenter Lab Deploy

Uses Terraform and Ansible to deploy a fully functional CVAD environment. Many of the scripts used are thanks to Dennis Span and his fantastic blog.

What it does

Deploys the following:

• 2 DDC Controllers with Director
• 2 Storefront Servers (Cluster)
• 1 SQL and License Server
• 1 Stand alone VDA

DDC

• Installs components including director
• Creates Citrix site
• Creates 1 Machine Catalog
• Creates 1 Delivery Group
• Creates 1 Published Desktop
• Creates 3 Applications
  • Notepad
  • Calculator
  • Paint
• Configures director
  • Adds logon domain
  • Sets default page
  • Removes SSL Warning

Storefront

• Installs Storefront components
• Creates Storefront cluster
• Configures Storefromt
  • Adds Citrix Gateway
  • Sets default page
  • Enables HTTP loopback for SSL offload
  • Adjusts logoff behavior

SQL and Citrix License

• Installs SQL and license server
• Installs SQL management tools
• Configures SQL for admins and service account
• Copies Citrix license files

VDA

• Installs VDA components
• Configures for DDCs

Prerequisites

• Need CVAD ISO contents copied to accessible share via Ansible account (eg \\mynas\isos\Citrix\Citrix_Virtual_Apps_and_Desktops_7_1906_2)
  • I used CVAD 1906 2 ISO
• Need SQL ISO contents copied to accessible share via Ansible account (eg \\mynas\isos\Microsoft\SQL\en_sql_server_2017_standard_x64_dvd_11294407)
  • I used SQL 2017 but other versions should work
• DHCP enabled network
• vCenter access and rights capable of deploying machines
• (optional for remote state) Terraform Cloud account created and API key for remote state.

Deploy machine

I used Ubuntu WSL to deploy from

1. Ansible installed
   • Install pywinrm pip install pywinrm and pip install pywinrm[credssp]
2. Terraform installed
3. Terraform-Inventory installed in path. This is used for the Ansible inventory
   • I copied to /usr/bin/
4. (If using remote state)Configure Access for the Terraform CLI
5. This REPO cloned down

vCenter Windows Server Template

1. I used Windows Server 2019 but I assume 2016 should also work.
2. WinRM needs to be configured and CredSSP enabled
   • Ansible provides a great script to enable quickly https://github.com/ansible/ansible/blob/devel/examples/scripts/ConfigureRemotingForAnsible.ps1
   • Run manually Enable-WSManCredSSP -Role Server -Force
3. I use linked clones to quickly deploy. In order for this to work the template needs to be converted to a VM with a single snapshot created.

Getting Started

Terraform

1. From the terraform directory copy lab.tfvars.sample to lab.tfvars
2. Adjust variables to reflect vCenter environment
3. Review main.tf and adjust any VM resources if needed
4. (If using remote cloud state) At the bottom of main.tf uncomment the terraform section and edit the organization and workspaces fields

terraform {
   backend "remote" {
     organization = "TechDrabble"
     workspaces {
       name = "cvad-lab"
     }
   }
}

5. run terraform init to install needed provider

Ansible

1. From the ansible directory copy vars.yml.sample to vars.yml
2. Adjust variables to reflect environment
3. If you want to license CVAD environment place generated license file in ansible\roles\license\files

Deploy

If you are comfortable with below process build.sh handles the below steps.

Note: If you prefer to run many of the tasks asynchronously switch the ansible-playbook lines within build.sh which will call a seperate playbook. This is faster but can consume more resources and less informative output.

#Sync
#ansible-playbook --inventory-file=/usr/bin/terraform-inventory ./ansible/playbook.yml -e @./ansible/vars.yml
#If you prefer to run most of the tasks async (can increase resources)
ansible-playbook --inventory-file=/usr/bin/terraform-inventory ./ansible/playbook-async.yml -e @./ansible/vars.yml

Terraform

1. From the terraform directory run terraform apply --var-file="lab.tfvars"
2. Verify the results and type yes to start the build

Ansible

1. From the root directory and the terraform deployment is completed run the following
   • export TF_STATE=./terraform used for the inventory script
   • Synchronous run (Serial tasks)
     • ansible-playbook --inventory-file=/usr/bin/terraform-inventory ./ansible/playbook.yml -e @./ansible/vars.yml to start the playbook
   • Asynchronous run (Parallel tasks)
     • ansible-playbook --inventory-file=/usr/bin/terraform-inventory ./ansible/playbook-async.yml -e @./ansible/vars.yml to start the playbook
   • Grab coffee

Destroy

If you are comfortable with below process destroy.sh handles the below steps. Please note this does not clean up the computer accounts

Terraform

1. From the terraform directory run terraform destroy --var-file="lab.tfvars"
2. Verify the results and type yes to destroy