Run: Cloud exit
Move off AWS, Google Cloud or VMware at your own pace
In short
Kafka.id moves companies off AWS, Google Cloud or VMware to infrastructure they control. We measure what you run and what it costs, then move it in stages to servers sized for real use, often with open-source virtualisation. Every step has a way back, and we only move what is cheaper or better to run outside.
Last updated
Who it is for
- Companies whose cloud bill grows faster than their business.
- VMware customers facing a much higher renewal.
- Teams that want their data on hardware they control.
What we do
- 1
Measure the bill
We break the bill down by service and find the big costs: idle servers, data transfer, managed databases and storage.
- 2
Plan the target
We design the new setup, for example rented dedicated servers running open-source virtualisation, and compare its full cost, our time included, with what you pay now.
- 3
Move in stages
We move one workload at a time, simplest first, and run old and new side by side until the new one has proven itself.
- 4
Run it
We keep the new infrastructure patched, monitored and backed up, or train your team to do it.
What you get
- A breakdown of what you run today and what each part costs
- A target design with a full cost comparison
- A migration plan with a way back for every step
- Workloads moved and tested before each switch
- Runbooks for the new setup
Questions
Is leaving the cloud always cheaper?
No. For small or very spiky workloads the public cloud is often the better deal. The savings come from steady workloads that run all day, heavy data transfer, or managed services priced well above the servers underneath. We do the sums before you decide.
What can replace VMware?
Open-source hypervisors such as Proxmox VE and KVM run the same kind of virtual machines, and most VMware machines can be converted. We test the conversion on copies of your machines before any switch.
How long does a move take?
It depends on how many systems you run and how tangled they are. A handful of servers can move in weeks. A large estate is a staged project over months. Every stage is planned and agreed before it starts.
What if something goes wrong during the move?
Every step has a way back. The old system keeps running until the new one has carried real traffic, and we take a snapshot before every switch.
Will we lose features like autoscaling?
Some. Autoscaling and managed services are what you pay the cloud for, and most steady workloads do not need them. Where one does, we keep that part in the cloud or build a simpler replacement.
Tell us what you're working on.
The brief takes about two minutes. Pick where to start:
Or email hello@kafka.id
