End of life is planned for XenApp & XenDesktop 7.15 LTSR in just over a year, on Aug 15, 2022. If your organization hasn’t started planning an upgrade to the next version of Citrix Virtual Apps and Desktops, you should. Upgrading your Citrix versions helps you to maintain supportability, stay on top of security fixes, and unlock new features to enhance the user experience.

But upgrades can be challenging because they require significant changes in the environment that fall outside the typical norms of day-to-day operations. Your Citrix upgrade doesn’t have to be a daunting experience, but it does require proper planning and change control to avoid disruption to your end users. In this blog post, I’ll provide an overview of the upgrade process and elaborate on in-place versus parallel upgrade methods.

In my role as Lead Enterprise Architect for Americas Customer Success, I get to speak with many customers, reviewing their current environment state, understanding business objectives, and mapping out how Citrix can bridge the gap to success. Through these interactions, I see a spectrum of customers, from those who have already completed their Citrix Cloud migration to those who run either an out-of-support version or an older LTSR version.

Which Version Should I Choose?

Over the years, Citrix has created version-servicing options to fit the needs of individual customers — Citrix Cloud, Current Release (CR), and Long Term Service Release (LTSR). Each differs in the length of support, features, and capabilities, as well as how fixes are introduced.

For customers focused on consistency, the LTSR version provides five years of mainstream support (with the option of five more years of extended support). Customers who need the latest features such as webcam optimization, EDT MTU Discovery, and reboot after drain enhancements choose the Current Release version, which requires more frequent updates. Citrix Cloud offers simpler management by facilitating access to the latest releases while giving control to the customer on the VDA workload version (including many public cloud integration enhancements).

The following diagram shows how these release cadences vary between the servicing options and how features cascade through.

To In-Place or Parallel Upgrade? That’s the Question.

I am often asked for recommendations on upgrading Citrix environments. With the advent of the FlexCast Management Architecture, in-place upgrades are easier. But that doesn’t mean that there isn’t risk associated with the process. First, some level of brokering interruption is required as the software undergoes its installation. For a Citrix Virtual Apps and Desktops controller, this would involve not just the installation of broker components but also a SQL database schema upgrade, as well. While existing user connections are not affected, it would prevent new connections for a short period, which is not acceptable for some organizations.

Additionally, the Citrix stack is composed of multiple components — License Server, Citrix StoreFront, Provisioning Services, and more — that have distinct upgrade requirements and associated complexity. The License Server, for example, could be upgraded in-place because there is a straightforward installation process (although I have seen a failed License Server upgrade take down a whole environment). Because each component has a different effect on the environment, you should have a plan for each component that may determine the in-place or parallel upgrade path.

Let’s look at the key differences among an in-place, a parallel upgrade strategy, and a Citrix Cloud migration.

In-Place Upgrade

The in-place upgrade process usually involves running the newer software installer on top of an existing server component. The benefit? It doesn’t require additional server infrastructure. Citrix has documented the full upgrade process here, and it’s shown in the flow chart below.

There are dependent components such as the License Server that need to be upgraded prior to the Citrix Virtual Apps and Desktops site. The Citrix Virtual Apps and Desktops site and PVS farm consist of client (i.e., VDA), server (i.e., controller), and database components (i.e., site, monitoring, and config logging DBs), which all are upgraded individually. While an in-place upgrade is appropriate for many, there are situations when the upgrade can fail and require a rollback. Citrix has also created a procedure for a non-disruptive Citrix Virtual Apps and Desktops upgrade that leverages the Local Host Cache feature to maintain brokering functionality while the SQL database is upgraded.

In-place upgrades are generally best when the process can be tested first in non-production environments that mimic the production environment. Alternatively, if you operate a pod architecture where there is a parallel set of passive infrastructure, user sessions can be flipped between sites to perform maintenance so the risks associated with an upgrade are lower. The passive site can be upgraded in place, and, after validation, users can be flipped to the upgraded infrastructure. Then, the upgrade process is repeated on the other pod. Some customers already have a well-documented and vetted process to upgrade in their environment for Cumulative Updates and/or Current Release versions. With this process, issues in the environment have been discovered and acknowledged, increasing confidence that an in-place upgrade could be smooth.

The advantages of this approach include:

Your considerations for this approach include:

  • In-place upgrades have a higher tendency for issues, and troubleshooting is typically more complex.
  • Testing an upgrade of an infrastructure component is difficult and production environment may be impacted.
  • Downtime planning is generally required.
  • Because VDAs are numerous, they can often take much longer than the infrastructure upgrade time. The order of operations is not strict here and Provisioning Services can be done, for example, after the site upgrade.

Parallel Upgrade

For parallel upgrades, a new infrastructure based on the new version is built separately from the existing component. After the new infrastructure is tested, production workloads can either be rebuilt or moved into this site. StoreFront Site Aggregation allows icons from multiple sites to be enumerated so users can go about their business with no change and the admin can manipulate resources to swing connections to the new site. The main benefit of this approach is the new environment begins from a clean state (i.e., less garbage in, garbage out) and workloads can be manipulated and migrated to the new site without users knowing.

Parallel upgrades are particularly appropriate for organizations looking to introduce other significant changes. Some examples include upgrading from legacy operating systems (such as Windows Server 2008 R2 to 2016/2019), moving or consolidating data centers, experiencing environmental issues requiring a fresh start, or delivering mission critical software requiring near-zero risk to unplanned downtime.

The advantages of this approach include:

  • The production environment is not affected by installation of new servers and components.
  • Testing is simpler and isolated.
  • It facilitates a more granular phased migration.
  •  A fresh build prevents unintentional configurations from being carried forward.

Your considerations for this approach include:

  • Additional server resources are required to build the new infrastructure in parallel.
  • Detailed planning and design work are required.
  • Machine Creation Services is tied to a particular site. Pooled catalogs can be re-built on the new site with relative ease because they’re based on a master image. Dedicated desktops require a procedure to remove desktops from the original site, re-register to the new site as an “existing” machine, and re-configure the assignment within the Delivery Group.
  • Multi-site aggregation can be a new feature to provide seamless switches between sites.

Citrix Cloud Migration

Adoption of Citrix Cloud is rapidly growing, and most conversations I have with customers now are about moving to Citrix Cloud because of the benefits it provides around security, user experience, and customer choice. This upgrade process is similar to the parallel upgrade process where Citrix Cloud provides a new Citrix Virtual Apps and Desktops service site and workloads can be transferred or rebuilt. A few benefits of this approach include:

  • Citrix Cloud and Citrix Virtual Apps and Desktops service alleviate the need for future infrastructure upgrades.
  • Citrix manages the control plane as a service.
  • The latest developments come to cloud first.
  • Simplifies and reduces overall Citrix management.
  • It’s a great way to connect into public cloud environments.

Customers can opt to combine the evergreen Citrix Cloud control plane with the stable LTSR VDA version and still get the 5+5 year LTSR support benefits. For healthcare customers in particular, this is a popular option because electronic health records systems can stay on a 1912 LTSR VDA version, for example, while Citrix Cloud allows flexibility for other major initiatives such as disaster recovery in a public cloud.

On the other hand, new use cases requiring the latest enhancements can be deployed using a Current Release version. VDA versions across workloads can be mixed where needed, while LTSR benefits would still apply to a subsection of the environment. To facilitate a simpler migration process, Citrix released the Automated Configuration Tool to simplify configuration machine catalogs, delivery groups, applications, and policies.

In March, I hosted a Citrix Healthcare Summit webinar on leading practices with cloud migrations where I talked through how Citrix Cloud can enable an agile hybrid-cloud solution, as well as a phased migration approach, as shown in the image below. For a deeper dive, check out the video.

The advantages of this approach are:

  • The upgrade process is similar to the parallel upgrade, as Citrix Cloud and Citrix Virtual Apps and Desktops are a separate site.
  • It aligns with future Citrix architecture and product direction.
  • It enables the latest features offered through cloud release cadence from the control plane perspective.
  • Customers can remain on an LTSR version of the VDA to support mission-critical workloads.

Your considerations for this approach are:

  • Citrix Cloud requires a new licensing model (although Citrix provides hybrid rights to facilitate a transition over a period of time).
  • The Automated Configuration Tool does not support every scenario, and in those events manual configuration is required
  • High availability and uptime are dependent on the Citrix Cloud service, and the Local Host Cache and upcoming Service Continuity features are recommended to augment availability.

So, Which One?

Deciding on which path to take will vary from organization to organization. An in-place upgrade typically affords the fastest path to supportability, although some level of brokering interruption is generally required. A parallel upgrade makes the most sense for organizations that are looking to modernize other parts of IT. Finally, Citrix Cloud introduces a new set of benefits and capabilities that enable an organization to reduce overall management, alleviate future infrastructure upgrades, and match a better path toward an overall cloud journey.

I encourage customers reach out to their Citrix representative, partner, and/or advisor with background knowledge on the environment to help you make an informed decision.

Disclaimer: The development, release and timing of any features or functionality described for our products remains at our sole discretion and are subject to change without notice or consultation. The information provided is for informational purposes only and is not a commitment, promise or legal obligation to deliver any material, code or functionality and should not be relied upon in making purchasing decisions or incorporated into any contract.