State agency rescues a failed Citrix migration and stabilizes its VDI

Site A ran an unstable Horizon deployment that crashed often, with session drops, desktop freezes, and inconsistent performance.

vSAN and Dell compute clusters at Site A were oversubscribed, misconfigured, and not monitored in a disciplined way.

Site B held a Horizon build abandoned mid-project, App Volumes packages that were incomplete or broken, and no standards for image lifecycle, patching, or configuration.

Citrix was still partially in production, which meant duplicate tool sets, extra complexity, and unnecessary licensing cost.

Active Directory held thousands of stale objects, Group Policy was layered and conflicting, and App Volumes and DEM were used without clear rules, causing random application failures.

DR was limited to basic backups with no regular validation, and the help desk carried a high volume of tickets on logons, profiles, printing, and session stability.

Z7 treated Site B as a greenfield reference design, deploying Horizon 8 components on supported, current versions, correcting vSAN policy, datastore layout, and host configuration, and expanding Dell compute to remove oversubscription. The team rebuilt App Volumes packaging and entitlements on a clean model, standardized DEM, and hardened and documented golden images with clear ownership and lifecycle.

With a clean reference in place, Z7 cleaned vCenter, snapshots, and orphaned virtual machines and fixed misaligned storage policies and cluster configurations. The team aligned App Volumes strategy and DEM configuration with Site B, stabilized performance, and removed recurring failure patterns.

In parallel, Z7 migrated the remaining Citrix hosted workloads into Horizon, consolidated and standardized golden images for the Horizon pools, and decommissioned Citrix components, dependencies, and licensing.

Z7 rebuilt the OU design to match business and technical needs, removed thousands of stale user and computer objects, and retired hundreds of obsolete or conflicting GPOs. Simplified logon policies led directly to faster and more reliable logons.

Z7 implemented two-site NetApp replication for key Horizon infrastructure and user data, used cloud for backup and DR validation without moving production desktops out of the data centers, and documented DR runbooks and tested failover patterns. Formal governance now covers image lifecycle, app packaging, and change control, backed by monitoring and analytics that trend recurring issues.

Improvement in login times after GPO cleanup and image tuning.

Fewer session crashes and disconnects.

Fewer recurring VDI support tickets, based on ticket analytics.

Citrix licensing left after the platform was retired.

Login times improved by about 48% to 60% after image tuning and GPO cleanup.

Session crashes and disconnects dropped by more than 80%, and users reported far fewer interruptions, especially during peak hours.

Overall VDI performance improved by an estimated 20% to 30%, largely due to vSAN remediation and added compute capacity.

Recurring VDI tickets fell by roughly 40% according to ticket system analytics, and support teams could shift more time from firefighting to planned maintenance and project work. Outages and emergency fixes decreased, which reduced soft costs and stress on staff.

Citrix licensing and support expenses were fully removed, and validated DR processes and cloud-supported backups lowered operational risk. Image, application, and policy ownership became clear and traceable.

Both data centers now run a single, well-governed Horizon 8 architecture on premises, the foundation for Phase 2: Horizon Cloud on Azure, FSLogix profiles, Turbo containers, and endpoint modernization.