Cloud exit · AWS and Azure to EU providers or your own servers

AWS or Azure to an EU provider — a bill you can predict.

Hyperscalers make sense for spiky, global workloads. Many company workloads are neither: the same VMs and databases running around the clock, billed by the hour, with fees on data that leaves. We move them to an EU hosting provider or your own servers on a fixed price and date, cut over outside business hours.

  • Fixed price & date
  • Cutover off-hours
  • Rollback at every wave
  • No lock-in

Why now

Leaving is cheaper than it used to be.

2024the year AWS, Microsoft Azure and Google Cloud started waiving data-transfer fees for customers who leave — on request
12 Jan 2027from this date the EU Data Act bans switching charges between cloud providers, egress fees included
EUproviders such as Hetzner, OVHcloud, IONOS or Scaleway — or your own racks; chosen by workload
0proprietary lock-in on the target: Proxmox, Kubernetes, PostgreSQL and S3-compatible storage run anywhere

Sources: AWS, Microsoft Azure and Google Cloud announcements on free data transfer for customers leaving (2024); Regulation (EU) 2023/2854 (Data Act), Art. 29.

How we move it

From hourly billing to a fixed monthly bill — without a big bang.

Cost and workload review

Which workloads are steady, which are spiky, and what they cost today — including data transfer, snapshots and support plans.

Target platform

Dedicated servers or VMs at an EU provider, or your own racks — Proxmox or Kubernetes, depending on what you run.

Managed services replaced

Managed databases → PostgreSQL or MySQL with replication; object storage → S3-compatible storage; load balancers → HAProxy or nginx.

Data transfer

Bulk copy ahead of time and a final delta sync at cutover; the data-transfer waiver is requested from your current provider.

Cutover

DNS switched in an off-hours window; the old environment stays until you confirm.

Operations

Monitoring, backups and patching from day one — by your team, by us or by StayPatched.

Straight talk

What doesn't move one-to-one

How a migration runs

Five steps. The old platform stays intact until you sign off.

  1. 1

    Assess

    Inventory, written plan, fixed quote and date — before any work starts.

  2. 2

    Pilot

    Non-critical systems move first; performance and restores proven on your workloads.

  3. 3

    Plan the cutover

    A runbook per wave, agreed windows, a written way back for every step.

  4. 4

    Cut over

    Wave by wave, outside your business hours. Anything wrong rolls back.

  5. 5

    Hand over

    Documentation, runbooks, full admin access — or StayPatched keeps it patched.

Questions

What people ask before this move.

Won't the data-transfer bill eat the savings?

Usually not: AWS, Azure and Google Cloud waive data-transfer fees for customers who leave, on request — and from 12 January 2027 the EU Data Act bans switching charges altogether.

Which EU provider do you use?

The one that fits your workload and compliance needs. We are not tied to any provider.

Do we lose managed databases?

You get PostgreSQL or MySQL with replication, backups and point-in-time recovery — run by your team, by us or by StayPatched.

Can we move part of it?

Yes. Moving the steady workloads and keeping the spiky ones in the cloud is often the best of both.

How long does it take?

It depends on the number of services and how deeply they use provider-specific features. The completion date is fixed in the quote after the assessment.

Get a migration plan

Tell us what you run. Get a fixed plan back.

A short call, then a written assessment: what moves as-is, what needs care, what should stay — with a fixed price and a fixed date. No obligation.

  • A reply from an engineer, not a sales sequence
  • Your details are used only to reply
  • If migrating isn't worth it for you, we'll say so

Your details are used only to reply to you. See the privacy notice.