Admin guide · Infrastructure

Your own runners

Run jobs inside your own network, next to the databases, file shares and systems behind your firewall. A runner connects out to the console over an encrypted connection; nothing has to reach into your network.

Where	Settings › Infrastructure › Containers
Who	An admin, or anyone whose role may enrol runners
Plans	On-premise runners: an add-on on Team, one cluster included on Business, included on Enterprise

The Containers page: counts of clusters, containers, running, no address and unreachable, and a card per cluster with New bootstrap token, Install script, Re-scan and a container table — Containers: one card per cluster, each with its containers, their status, and the buttons to add and update them. Click to open full size.

Register your own runners

Press + Add container and name the cluster for the environment it serves. The cluster comes first, so a bootstrap token can be issued for it.
On the cluster's card, press New bootstrap token, and copy it.
Press Install script and run it on the machine that will host the runner, with the token. The runner enrols with its own key and connects out to the console.
The container appears in the cluster's table. Status turns to online when it answers; Deploy target says whether jobs may run there.
Press Doctor on a container to check it end to end.

Read the cluster cards

Column	What it says
Container	The runner's name
Status	Online, offline, or no answer
Address	Where the console reaches it, or "over the hub" when it calls out
Discovered	How the console found it
Last seen	Its last heartbeat
Deploy target	Whether jobs may be deployed to it
Doctor	A check of the container end to end

Keep them up to date

When a container is behind the newest version, its card says so - "run the install script again to update". Run it again on that machine; Re-scan refreshes the card. Containers are moved to a new version one at a time.

How healthy each runner is, over days: Runtime health.

Last updated October 29, 2026