Operators

Guides for the people who run Nimbus: DevOps, platform, and admin teams self-hosting a server for their organization — or developers hosting their own. Start with the self-host quickstart if you haven’t run a server yet, then work through deploying to a Linux server.

Install & deploy

• Deploy to a Linux server — the end-to-end tutorial: install, systemd unit, token, first request.
• Container image — run the published OCI image with podman or docker.
• Desktop install — the CLI and desktop app on a workstation.
• Updates — keep the binary current.

Data

• Storage backends — SQLite, redb, Postgres, MySQL, and libSQL topologies.
• Encryption at rest — key providers, migration, and rotation.
• Backup and restore — nimbus backup for the embedded providers, plus what to capture per backend and how to restore it.
• Object storage — administer the object byte plane: placement policy, the master key, GC and erasure health, byte-plane backup and restore, and destructive tenant removal.

Administration

• Manage tenants — create and administer tenants on a running server.
• Run Nimbus as a service — systemd and Quadlet service management with nimbus node.
• Scale out — grow past one machine by partitioning tenants across independent instances.
• Inspect the server — health, debug endpoints, logs, and the access audit log.
• Security hardening — the checklist for exposed deployments.
• Troubleshooting — symptom, cause, fix.

When one machine runs out of headroom, scale out by partitioning tenants across instances; the single-process growth model is explained in Concepts: scaling. The flag-by-flag configuration tables live in the Reference section, and the isolation model itself is explained in Concepts: tenant isolation.