Study guide
Technical reference and lesson notes
Citrix Provisioning Services, commonly referred to as Citrix PVS, is one of the most powerful image management technologies in the Citrix stack. Instead of building, maintaining, and patching hundreds of individual virtual machines with their own full operating system disks, PVS allows administrators to create a single master image and stream that image to many target devices over the network.
The core idea is simple but extremely powerful. You build one clean master machine, capture its disk into a vDisk, and then allow many virtual machines to boot from that shared vDisk. This makes large-scale desktop management much easier, especially in environments with hundreds of Windows desktops.
Why Citrix PVS Exists
In a traditional virtual desktop environment, every virtual machine has its own hard disk. If an organization needs 300 Windows 10 or Windows 11 desktops, that could mean 300 separate virtual disks to store, patch, manage, and protect.
That approach works, but it creates storage overhead and operational complexity. Every VM consumes storage, every image needs to be maintained, and every change can become time consuming.
Citrix PVS solves this by allowing many target devices to share the same base image. Instead of each VM having a full operating system disk, the machines boot across the network from a centrally managed vDisk.
This is especially useful for environments such as training labs, call centers, task worker desktops, shared desktop pools, and any situation where many users need the same standardized Windows desktop experience.
The Basic PVS Architecture
A typical Citrix PVS deployment starts with one or more PVS servers. In production, it is common to deploy at least two PVS servers for redundancy. If one PVS server is unavailable, another PVS server can continue streaming the vDisk to target devices.
The PVS servers are usually installed on Windows Server and connected to shared or replicated storage where the vDisk files are stored. The storage layer is important because the vDisk is the source image that target devices will boot from.
There are a few common ways to handle the vDisk storage. Some environments use shared storage such as a CIFS share. Others use fast local disks on each PVS server and replicate the vDisk files between servers using a tool such as Robocopy or another replication method. The important point is that each PVS server needs access to the same vDisk content so that streaming remains consistent and highly available.
Creating the Master Image
The process starts with a master virtual machine. This is usually a normal Windows desktop VM that is built, patched, configured, and prepared exactly the way the organization wants the user desktops to look.
On this master machine, the Citrix VDA is installed so the machine can register with the Citrix Delivery Controller and participate in a Citrix Virtual Apps and Desktops environment.
Once the master VM is ready, the PVS imaging tools are used to capture the local hard disk and convert it into a vDisk file. This process takes the disk from the master VM and stores it on the PVS storage location as a vDisk, commonly using a format such as VHD or VHDX.
At that point, the master disk has effectively been turned into a centrally managed boot image.
The vDisk in Standard Mode
One of the key concepts in PVS is the vDisk mode. For pooled desktops, the vDisk is commonly placed into Standard Mode.
Standard Mode means that the vDisk is read-only and can be shared by multiple target devices at the same time. This is where the real value of PVS comes in. Instead of each virtual desktop having its own full copy of Windows, many desktops can boot from the same shared read-only vDisk.
Because the vDisk is read-only, changes made during a user session are written somewhere else, typically to a write cache location. This allows the base image to remain clean and consistent while still allowing each running VM to function normally.
Creating Diskless Target Devices
After the vDisk is created, the next step is to prepare the virtual machine template. In a PVS design, the target devices do not need traditional operating system disks.
A common process is to clone the master VM or create a template from it, then remove the virtual hard disk. This results in a diskless VM template. The VM still has its virtual hardware, CPU, memory, network adapter, and other settings, but it no longer boots from a local disk.
Instead, it will boot from the network and receive its operating system image from PVS.
This is one of the major differences between PVS and a traditional manually built VM environment. The target machines exist as virtual machines, but their operating system is streamed to them from the PVS server.
Using the PVS Console to Create a Catalog
The PVS console includes wizards that help create and organize target devices. During the catalog creation process, PVS needs to communicate with the Citrix Delivery Controller and the hypervisor.
The wizard typically asks for information about the Citrix broker, the hypervisor connection, the VM template, the vDisk to assign, and the target device collection.
A PVS collection is essentially a group of target devices. Each target device has a name and a MAC address, and PVS uses that information to determine which device is booting and which vDisk it should receive.
For example, if an administrator needs 300 Windows desktops for a group of users, PVS can create those target devices, associate them with the correct vDisk, and help create the required machine accounts in Active Directory.
How the Network Boot Process Works
The PVS boot process is where the technology becomes really interesting.
When a diskless target device powers on, it performs a network boot. The device uses DHCP to receive network configuration and boot information. In many environments, DHCP options 66 and 67 are used to point the target device to the PVS boot server and bootstrap file.
The bootstrap file is delivered using TFTP. Once the target device receives the bootstrap, it contacts the PVS server and asks whether its MAC address exists in the PVS database.
If the MAC address is known, PVS identifies the correct vDisk assigned to that target device. The PVS server then begins streaming the operating system blocks that the target device needs.
The target device does not download the entire vDisk all at once. Instead, PVS streams the required disk blocks as needed. Conceptually, it is similar to streaming video, except instead of streaming media content, PVS is streaming the operating system disk to the target machine.
Why Network Performance Matters
Because PVS streams the operating system over the network, network performance is critical. A stable and properly designed network is required for a successful PVS deployment.
At minimum, a 1 Gb network is commonly expected, although larger environments may benefit from faster networking and additional optimization. PVS can perform extremely well when the network, storage, and server design are properly planned.
The design should account for redundancy, storage performance, network throughput, write cache placement, and the number of target devices that may boot at the same time.
PVS and Citrix Delivery Groups
After the PVS target devices are created and available in Citrix Studio, administrators create or update the appropriate machine catalog and delivery group.
The delivery group controls which users can access the desktops. Once everything is assigned correctly, users connect through Citrix as usual. From the user’s perspective, they are simply launching a desktop. Behind the scenes, that desktop may be a diskless VM booted from a shared PVS vDisk.
Power management can also be used to start machines before users need them. For example, if 150 users begin work in the morning, Citrix can power on a pool of desktops so they are ready when users log in.
Why Citrix PVS Is So Powerful
Citrix PVS is powerful because it centralizes image management. Instead of managing hundreds of separate virtual disks, administrators manage one primary image.
When the image needs to be updated, the administrator updates the master image or creates a new vDisk version, tests it, and then promotes it when ready. This makes patching, application updates, and desktop consistency much easier to control.
PVS is especially valuable when an organization needs a large number of standardized desktops and wants to reduce storage consumption, improve consistency, and simplify image maintenance.
Final Thoughts
Citrix Provisioning Services is one of the most impressive technologies in the Citrix platform because it changes the way virtual desktops are delivered. Instead of treating every desktop as a fully independent machine with its own operating system disk, PVS allows many machines to boot from a shared, centrally managed vDisk.
The key components are the PVS servers, shared or replicated vDisk storage, the master image, the vDisk, target device collections, DHCP/TFTP boot services, and the Citrix machine catalog and delivery group.
Once these pieces are in place, PVS can stream a single Windows image to hundreds of virtual desktops. For environments that need scale, consistency, and efficient image management, Citrix PVS remains a very strong solution.