Virtual desktop platforms
Running the Citrix and Omnissa platforms of seven clients, from 50 to 300 users each, and preparing and carrying out their upgrades.
- Case study
- 01 / 07
- Projects
- Citrix · Omnissa · VMware · Azure
- Type
- Professional project (employment)
- Role
- Contribution to a team project (Virtual Desktop team)
- Period
- 06/2024 – 12/2025
- Status
- Completed
01
Overview
From June 2024 to December 2025, I was a systems administrator in Evok Altern8's Virtual Desktop team in Fribourg. I ran the Citrix Virtual Apps and Desktops and Omnissa Horizon platforms of seven clients, from 50 to 300 users each, and independently handled a portfolio of fifteen clients.
This case study covers five parts of that job: Citrix upgrades, VMware and disaster recovery, the move from Citrix to Parallels RAS, FSLogix profiles, and level-2/3 support.
02
Context
On these platforms, users work on remotely published desktops and applications: when the platform stops, their work stops too.
03
Challenge
Upgrading production platforms without interrupting their users' work.
04
Objectives
- Keep the platforms in working order
- Handle level-2/3 incidents, major ones included
- Upgrade Citrix
- Hand procedures over to level-1/2 teams
05
Architecture
Not detailed here: these are client environments.
06
Technologies
- Citrix Virtual Apps and Desktops (VDA, Delivery Controllers)
- StoreFront
- FSLogix
- NetScaler Gateway
- Omnissa Horizon
- Parallels RAS
- VMware vSphere / vCenter / ESXi
- Windows Server
- Active Directory (GPO)
- DFS
- AppLocker
- SharePoint
- Entra ID
- Microsoft 365
- Intune
- Exchange
- Microsoft Azure (IaaS, PaaS, Azure Virtual Desktop)
- Log Analytics
07
Implementation
Citrix upgrades. I prepared and carried out the upgrades of the VDAs, of the Delivery Controllers (update, configuration, checks) and of the virtual machines in vCenter, on scopes of 10 to 100 users — for example three host servers and the master server. With fewer than three VDAs, I updated them by hand; beyond that, I rolled the update out gradually, in groups over several days, with a script prepared by the Account Manager: I knew how to use it and could suggest changes to it, but I did not write it. Before each intervention I took a snapshot, kept for 48 hours. Before and after, I checked a test account, the DFS permissions, the printers and the launch of the published applications.
VMware and disaster recovery. On an infrastructure of four ESXi hosts — two for production, one for internal tools and the DMZ, one dedicated to disaster recovery — I took part in managing vSphere and the VMware licences, managed virtual machines and their resources (checks, and increases following contract amendments), took snapshots and ran restarts. I took part in a disaster-recovery simulation: after a prior audit of the infrastructure against a checklist, outside production hours and with the client's agreement, the production hosts were shut down from vCenter and several scenarios were tested, with different combinations of virtual machines powered on and off. I checked the machines, took part in the return to normal, and wrote the recovery documentation for the client.
From Citrix to Parallels RAS. I installed and configured Parallels RAS from the existing environment: published applications recreated, user groups and access rights carried over, existing user profiles kept with FSLogix, and tests before users were switched over, in which I took part. A colleague had prepared the directories and the documentation, based on Parallels' official resources.
FSLogix. FSLogix was already deployed. I troubleshot locked user profiles, VHDX file mounts and persistent sessions, and worked in environments using AppLocker.
Level-2/3 support. I handled access and permission issues on SharePoint, DFS permissions, the checking and troubleshooting of GPOs, Microsoft 365 support, and VDI incidents affecting users and published applications.
I also deployed Windows Server machines, and administered Microsoft Azure resources (IaaS, PaaS, Azure Virtual Desktop), with Entra ID, Conditional Access and MFA.
08
Results
40 to 50 level-2/3 tickets handled per month, from June 2024 to December 2025. Lessons-learned reports, post-mortems and procedures written for level-1/2 teams, and recovery documentation written for a client.
09
My part
Running the platforms of seven clients and Azure resources, and an independent portfolio of fifteen clients. I prepared and carried out the Citrix upgrades on my own on the technical side, with the Account Manager available remotely for supervision and escalation. I installed and configured Parallels RAS, troubleshot FSLogix and handled level-2/3 support.
Not mine: the VDA deployment script (prepared by the Account Manager), the Parallels RAS directories and documentation (prepared by a colleague), and the design and initial deployment of FSLogix.
10
Lessons learned
On a virtual desktop platform, service continuity depends on consistent day-to-day operations.