Skip to content
BigFleet Providers

Azure provider

The Azure provider gives your BigFleet fleet machines to run on. When BigFleet decides your clusters need more capacity, the provider creates Azure Virtual Machines; when the fleet scales in, it drains and deletes them. You point it at your subscription, a resource group, and a subnet, and it provisions pay-as-you-go and Spot capacity automatically — no manual VM management, no VM scale-set babysitting.

You run one process per region (location), next to BigFleet. Each process owns a single region’s capacity, and BigFleet dials it to request, configure, drain, and delete machines as demand moves.

How it behaves

  • Hardened and operable. It ships as a container image and a Helm chart, runs non-root on a distroless, read-only root filesystem, and exposes liveness/readiness probes, Prometheus metrics, and structured logs. See Observability.
  • Certified. It passes the full BigFleet provider conformance program — 93 certified behaviors — credential-free on every change, plus an extension suite that asserts stronger invariants. See Certification.
  • Conservative by default. A Create returns only once the VM is actually provisioned, Spot machines always carry a real interruption risk (never a falsely-cheap zero), and a failed bootstrap or drain surfaces as a hard failure rather than a silently-broken node. Capacity it doesn’t own, it never touches.

What you need

To run it against a real region, have these ready (the Credentials page walks through the identity):

  • An Azure subscription and the region you want capacity in (one process per region).
  • A resource group the provider creates VMs in, and a VNet/subnet for it to attach NICs to.
  • A base image that joins your cluster (an image URN like Canonical:ubuntu-24_04-lts:server:latest, or your own managed image). The provider creates the VM, then delivers a per-cluster bootstrap blob via a CustomScript extension and runs a small hook your image ships. The hook contract is in Configuration.
  • A managed identity with a least-privilege role scoped to the resource group, federated to the chart’s ServiceAccount via Workload Identity on AKS — ready-to-apply Terraform is on the Credentials page.

Deploy it

The provider is a published container image plus a Helm chart — you don’t build from source. The path on AKS is:

  1. Create the managed identity with a role scoped to your resource group, and federate it to the chart’s ServiceAccount. The Credentials page has the exact role and Terraform.
  2. Install the Helm chart, one release per region, pointing it at your location, subscription, resource group, subnet, image, and your offerings (the quota of capacity it may provision). Enable durable state on a PersistentVolume so bindings survive restarts.

A minimal install is on the Install & deploy page.