Servers approaching end of life, storage struggling to keep up with growth, changing licensing costs, or an environment that has become too complex to manage: sooner or later, these situations lead companies to consider how to update their IT infrastructure.
The most immediate option is to replace each component with its newer equivalent. But if the limitations depend on the architecture rather than simply on the age of the hardware, there is a risk of investing only to face the same problems again a few years later.
That is why it is better to start with a different question: what will the infrastructure need to support over the next few years, and which constraints of the current environment are not worth carrying forward?
Why update your IT infrastructure?
Hardware end of life is one of the most obvious reasons. Support, warranty coverage, spare-part availability, and compatibility with new software versions may make an upgrade necessary.
But it is not the only sign. An infrastructure update may also become appropriate when workloads and data volumes increase, business continuity requirements become more demanding, growth becomes difficult to manage, or operating costs and complexity are no longer aligned with the company’s needs.
It is therefore useful to distinguish between two issues: the obsolescence of individual components and the suitability of the architecture as a whole. If the limitation is structural, replacing only the servers may simply postpone the problem rather than solve it.
How to plan the upgrade: start with real workloads
Before defining the new hardware configuration, you need data about the existing environment.
CPU, RAM, and storage are the starting point, but they must be evaluated based on actual usage. The resources allocated to VMs may differ significantly from those they actually use. For storage, knowing how many terabytes are occupied is not enough: I/O, latency, data growth, and the required level of redundancy also matter.
The analysis should consider:
- average and peak CPU and RAM usage;
- storage capacity in use and expected growth;
- workload characteristics and I/O requirements;
- VMs and services that are truly critical;
- resources required in the expected failure scenario;
- new projects and workloads expected over the next few years.
The goal is not to size the new system based on today’s average usage, but to find the right balance between actual usage, peak demand, resilience, and future growth.
Upgrade or modernize the infrastructure?
Not every upgrade requires a change in architecture.
If the environment works well and the main issue is hardware age, renewing the existing platform may be the right choice. If, on the other hand, systems, storage, and resources have become fragmented over the years, the refresh can be an opportunity to consolidate or evaluate a different model.
This is where, for example, a hyperconverged infrastructure may come into play, with compute capacity and distributed storage designed as part of the same cluster.
However, this is not an automatic choice. The number of nodes, network architecture, storage performance, latency, and operational expertise must all be appropriate for the workloads. The goal is not to eliminate the SAN at all costs, but to understand which architecture is most sustainable for the actual environment.
Here you can find the AXOL systems engineers’ in-depth guide to hyperconverged systems.
Migration and business continuity are part of the project
Updating the infrastructure does not simply mean installing new hardware. You also need to determine how services will be transferred to the new environment.
Which VMs can be migrated easily? Which applications require a scheduled maintenance window? How much downtime is acceptable? Have the backups been verified before migration?
These considerations can affect the architecture, timeline, and overall cost. For this reason, the target architecture and migration plan should be designed together.
The financial comparison should also go beyond the price of the hardware and take into account configuration, migration, backup, any subscriptions or licenses, support, and ongoing management.
What to do with the existing hardware
Updating does not necessarily mean decommissioning everything. Servers that are still reliable can take on a secondary role, for example in backup or disaster recovery, after their condition, performance, power consumption, compatibility, and supportability have been assessed.
In the Union Gas e Luce project, some of the previous servers were repurposed for Proxmox Backup Server and disaster recovery. The value lies not in reuse itself, but in assigning existing hardware a role that is consistent with the new project.
Read the Union Gas e Luce case study.
Two AXOL case studies, two different reasons to upgrade
Ginvest started with a traditional server infrastructure that was no longer adequate for the growth of its digital services, document volumes, and business continuity requirements. This was compounded by the end of the vendor’s software support and the difficulty of sourcing hardware components. Read the full Ginvest case study.
Union Gas e Luce, on the other hand, was already using Proxmox VE, but on a heterogeneous infrastructure that had grown over time and relied on shared SAN storage. In this case, the upgrade was not driven by the need to change hypervisors, but by the need to make the architecture more scalable, flexible, and easier to manage.
Two different situations that lead to the same principle: the new configuration should be the result of the analysis, not the starting point.
Where to start when updating your IT infrastructure
Before requesting a new configuration, collect the essential data about the current environment: VMs, actual CPU and RAM usage, storage capacity and growth, I/O, business continuity requirements, backups, application constraints, and workloads expected over the next few years.
From there, it becomes possible to determine whether it makes more sense to upgrade the current architecture, consolidate resources, or adopt a different model.
Do you need to update your servers, storage, or virtualization platform? Get an initial configuration proposal based on the available data and review it with AXOL specialists before moving on to the final technical proposal.
