Open source is a promise, not a feature.
Most providers run open source and lock you in anyway. This page explains how that happens, and what we do differently.
The lock-in most people miss
A provider can run entirely open source underneath and still be extremely difficult to leave. The software licence isn't what traps you. These are:
A proprietary control plane.
The hypervisor is open. The thing you actually operate it with is not, and it only exists here.
Extended APIs.
Standard endpoints, plus a handful of provider-specific ones that your automation quietly comes to depend on.
Portal-only automation.
Everything is configurable, as long as you're clicking. Nothing is expressible as code you could run elsewhere.
Undocumented configuration.
The cluster works. Nobody outside the provider knows precisely how it was built, so nobody else can take it over.
Egress priced as a deterrent.
Moving 200 TB costs enough to make staying the rational decision, independent of whether staying is the right one.
No defined exit.
Nobody wrote down what leaving involves, so leaving becomes a discovery project you have to fund before you can even scope it.
None of this requires bad intent. It's what happens by default when a provider optimises for its own operations. But the effect on you is identical either way.
How we're built differently
Choose the stack, not the vendor
OpenStack, Proxmox, Ceph, Kubernetes — we run all of them, so we have no reason to push you toward one. Sometimes the answer is that you don't need a private cloud at all, and a handful of bare metal machines will do. We'll tell you that too, and it costs us the larger deal.
Placement follows the workload
Residency, latency and compliance decide where your infrastructure runs — including your own facility, if that's the right answer.
Predictable economics
Fixed monthly cost, quoted up front. Scaling you can model in a spreadsheet before you commit to it. No per-API-call billing, no egress surprises, and no line items you need a specialist to interpret.
The four claims
No vendor dependency
No geographic dependency
Stack fit, not stack religion
Predictable economics
What we're not
We’re not a hyperscaler. We don’t have thirty regions, we don’t have two hundred managed services, and if you need a global CDN with a hundred edge locations, we’re not it.
What we do is deliberately narrow: dedicated open infrastructure, run by senior engineers, that you could leave if you wanted to. Everything we’ve built serves that, and we’ve turned down work that didn’t.
If that’s not what you need, we’ll say so on the first call.
Proof
The same documentation that tells a customer how to run on Racktrail contains the steps to leave it. Rather than quote someone saying we're portable, here is the page itself.
-
Export images
qcow2 / raw
-
Retrieve state
Terraform + Ansible
-
Replicate storage
Ceph / S3 sync
-
Cut over and close
documented in the same runbook
Tell us what needs to run.
Send us the workload, the constraints, and the region it has to live in. An engineer — not a sales rep — will tell you honestly whether we're the right fit.