Migration to cloud and Kubernetes in stages, with a documented way back.
Servers nobody dares to change, ever more expensive licences and a plan that never closes. We plan the migration to cloud and Kubernetes in stages, with a documented way back and the point of no return defined before we start.
01 · Situation
Self-check
Tick the situations that apply to your company. Your selection stays in your browser: no data is sent.
Tick those that apply to your company.
Schedule a call02 · Why now
Dates worth keeping in mind
Public dates from the vendors and the EU Data Act. Follow each source to check them.
End of general support for vSphere 7
If your virtualisation is still on version 7, it no longer has general vendor support. It is time to upgrade, renegotiate or migrate.
Source: vSphere 7, End of General Support · 2 Oct 2025End of extended support for Windows Server 2016
No standard security patches. Migrating with an inventory, testing and a change window isn’t something you do in an afternoon.
Source: Microsoft Lifecycle · Jan 2027Switching cloud provider, with no switching charges
Under the EU Data Act, from that date providers can no longer charge for switching or for moving your data out. Negotiating or moving gets easier, if you know what you have.
Source: Reglamento (UE) 2023/2854, Data Act · 12 Jan 2027
03 · What changes
What changes when the migration has a plan
Today: Everything at once, one weekend.
With Nubyron: In stages, with acceptance criteria for each.
Today: If it goes wrong, you improvise.
With Nubyron: A documented way back and an agreed point of no return.
Today: Nobody knows what depends on what.
With Nubyron: A dependency map and a reasoned decision per application.
Today: The knowledge leaves with the provider.
With Nubyron: Documentation and handover to your team.
04 · What you receive
What we deliver
A fixed project, in stages and in writing. This is what you take away:
- A plan in stages
With acceptance criteria, owners and a window for each change. Nothing moves without knowing how to go back.
- Inventory and dependency map
What you have, what depends on what and a reasoned decision for each application.
- A documented way back
And the point of no return set in advance, with your sign-off.
- Target tested in your accounts
On AWS, Azure or Google Cloud, and on Kubernetes when it pays off.
- Documentation and handover
So your team runs what was migrated without depending on us.
05 · How we work
How we work, step by step
- Assessment
We inventory what you have and what each piece depends on, and decide what to do with each application.
- Destination and plan
We prepare the destination (a minimal landing zone to migrate into) and the staged plan.
- Move and validate
We migrate in batches, starting small, and test before the real change against agreed criteria.
- Rollback and handover
The rollback is written down up to the point of no return. We do not switch off the source the same day, and we hand over to your team.
06 · Other services
Complementary services
07 · Further reading
Recommended reading
- ARQUITECTURA CLOUDReference architecture: PDF processing on Azure
- KUBERNETES & SREKubernetes shouldn't give you mini-heart attacks
08 · FAQ
Frequently asked questions
Will there be downtime?
It depends on each application. We see it in the inventory and every change has a window agreed with you. We do not promise “zero downtime” without having measured it.
Do we have to move everything to Kubernetes?
No. Only when it pays off. For many applications a virtual machine or a managed service is the better option, and we tell you so.
And if we decide to stop halfway?
That is what the staged plan is for: each stage ends in a stable, documented state, with a way back up to the point of no return.
How do you migrate with the least risk to the business?
You assess before moving anything, migrate in stages with acceptance criteria, and keep the rollback available up to a point of no return marked in advance. If there will be downtime, it is agreed in the plan; we do not promise zero interruptions.
When does migrating to Kubernetes pay off?
When you have several services, frequent releases and a team able to operate it. If your application is simple, a simpler managed service may be better. We assess it with you and do not recommend it by default.
What is the point of no return?
The moment after which going back is no longer viable, mostly because of data. We mark it in advance, together with who decides and which signal forces a return.
No 24x7 on-call or SLA · No zero-risk claims · No guaranteed savings · No testing without permission. See everything we do not promise →
Shall we talk about cloud and Kubernetes migration?
30 minutes with a technical member of the team. We review what you ticked, tell you where we would start and what can wait. No preparation is needed.
First step
Tell us about your case
Write to us about your environment and your question in general terms. We propose a written scope before any work begins.
- A technical person repliesYour message is reviewed by a technical specialist.
- Scope in writing before we startWhat is included, what is not and what it costs. You decide with the proposal in writing.
- Your data, only to reply to youNo advertising unless you agree.
Prefer to talk?
Talk to us for 30 minutes