Upgrading an outdated client management infrastructure—parts of which were no longer functional—at 15 locations, followed by ongoing operation and a structured handover.
Background
The client management infrastructure of an industrial company with locations distributed worldwide had grown over the years and was, in many key areas, no longer functional.
Specifically, this meant that application downloads would abort without any discernible pattern. It was not possible to distribute patches. There was no centralized, standardized operating system installation. And not all locations were even connected to the central management system; at some, local workarounds had become established.
The state of the documentation only made matters worse. There was no reliable overview of which servers performed which tasks, how applications were distributed, or which rules were actually in effect. Before any decision could be made, the system’s configuration first had to be reconstructed.
This meant the system was not just outdated. Without a reliable picture of the current state of affairs, it is impossible to either operate or modernize the system.
Solution
Therefore, the first step was not the setup, but rather an assessment of the existing system. Over the course of several months, we identified which components existed, what functions they served, and how the various locations were connected. This phase was the real challenge of the project and laid the groundwork for everything that followed.
Based on this, the new infrastructure was implemented: additional management servers and distribution points located worldwide, a cloud management gateway for devices outside the corporate network, and local PXE booting at the sites. Communication between clients and servers was switched to certificate-based encryption. Sites that had previously been left out were connected.
Building on this foundation, the following operational processes were developed: a multi-stage update management system, a complete operating system installation for clients and servers via MECM, a structured driver management system, and compliance rules that ensure corporate policies are technically enforced.
Approximately 300 applications were migrated to MECM and prepared for fully automated installation, along with 15 to 20 update packages per month. The migration was carried out using the ceterion Packaging Framework and its MECM import module. Among these were about 30 CAD applications whose versions and configurations are specified by the client and therefore allow no flexibility in the packaging process.
The Citrix Golden Master and the images for Omnissa Horizon are also created using the same infrastructure. This provides a single method for distributing all software, rather than separate processes for each platform.
Operation and Handover
Once the setup was complete, ceterion took over operations. The environment was handed over to the client’s new operational organization as part of a structured transition. This included comprehensive operational documentation, knowledge transfer, and a guided handover phase.
Outcome
- 2,500 clients and 350 servers at 15 locations worldwide, managed centrally
- Approximately 300 applications can be installed fully automatically, plus 15 to 20 update packages per month
New installations, updates, and application deployment are carried out globally through a standardized process. Company policies are no longer communicated via instructions but are technically enforced through compliance rules and are therefore verifiable. And the environment is documented in such a way that another team could take it over. That was the situation that was missing at the beginning.