Configuration
Daemon config.toml, controller environment variables and install paths.
Observability
Prometheus metrics, OTLP export, Linux pressure signals and fleet-wide scrapes.
Security
Authentication, RBAC, audit, TLS and how to report vulnerabilities.
Production hardening
The production security profile, secret and key lifecycle, audit retention and the mandatory eBPF gate.
Upgrade and backup
Upgrade Machina in place, back up VMs and controller state, and roll back.
Troubleshooting
The first checks to run, and fixes for the most common problems.
Bare-metal troubleshooting
Symptoms, causes and fixes for bare-metal provisioning.
Failure domains and rack loss
Label hosts with a site and rack, spread anti-affinity groups across racks, and check whether HA could restart a lost rack.
Air-gapped and sovereign sites
Run Machina with no route to the internet - what to mirror, which settings point at your mirror, and how to prove nothing leaves the site.
vGPU partitions
Split a GPU into mediated-device partitions and give each to a VM, instead of passing the whole card through.
Secure Boot and vTPM
Create VMs with UEFI Secure Boot enforced and an emulated TPM 2.0, the combination Windows 11 and hardened Linux images expect.
Fleet Kubernetes clusters (k3s)
Create, scale, upgrade and delete k3s clusters across enrolled hosts from the controller.
Desktop pools (simple VDI)
Clone a template VM into a pool of desktops, hand one to each user who checks out, and recycle it when they are done.
Off-site backup replication
Copy backups to an S3-compatible store at another site, verify them, and pull one back for a restore.
iSCSI storage
Log in to an iSCSI target, expose its LUNs as a storage pool, and attach a LUN to a VM as a block disk.