FAQ
Questions teams ask before joining the pilot
If you are evaluating Infraspawn for private infrastructure, on-prem deployments, or mixed-node environments, these are the most common questions so far.
What can I deploy with Infraspawn?
Anything that runs on Linux, natively or in Podman and Kubernetes. The catalogue currently leans towards what IoT and building systems need around them: MQTT, DHCP, timeseries storage, databases, reverse proxies, alongside internal tooling like file servers, Git hosting, and local LLM stacks. It grows with the pilots we run, and your own images run alongside it.
What if a service is not in the catalogue?
You are not blocked on us. If you have a container build, you can run it today: add the container registry service to a machine, push your image to it, then add the custom image service wherever it should run and point it at that image. It joins the network, gets connected to what it depends on, and is documented like anything else. Tell us what you would rather have in the catalogue and we will prioritize it, since the pilots we run are what decides the order.
Do I need Docker/Podman or Kubernetes expertise?
No. The networking, the policies, the images and containers, and the deployment itself are handled for you. Once a service is assigned you can read the container setup Infraspawn generated for it: the work you would otherwise have hired someone to do, done and written down.
Do I need Tailscale or another VPN alongside this?
No. The mesh is part of Infraspawn, not something you have to bring: traffic between machines is encrypted, and they make the connection themselves even when both ends sit behind a home router or a company firewall. Where that is not possible we relay the traffic and tell the admin. There is no second vendor charging you for the same machines, and you keep sight of where your traffic goes: you choose the meeting point machines connect through, once for the account and again per network where it matters. Running those on your own servers is on the roadmap for enterprise customers.
How does a machine get onto a network?
The machine is created in Infraspawn first. Create it in the dashboard, with isc, or in a file you apply, and you get back a key for that machine and no other. Install the agent on the real hardware, give it that key, and it comes up as the machine you created. A network is never left open for machines to ask their way in, so there is no access list to write and review: a network holds what you put in it and nothing else.
How fast can I be up and running?
If the services you need are in the catalogue, minutes. Install the agent on your machines, hand them a node key, and the control plane takes it from there.
This is too much black box magic for my taste.
Then read it. The documentation comes from the same source as the setup each machine runs, so what is deployed and how it is configured is written down rather than left in the head of whoever set it up. The file you applied stays the record of what you asked for, and a dry run tells you what a change would do before it does it.
I want to sign up. When will this be self-service?
Infraspawn is available through private pilots right now. We work directly with the teams deploying into real customer environments, and what those deployments need is what decides the catalogue and the order it gets built in. Self-service follows once that catalogue covers enough of what pilots ask for. Joining one is how you get a say in it.