Kopiur

Kopiur, from Kopia plus Rust, is a Kopia-native Kubernetes backup operator written in Rust on kube-rs.

It makes a kopia repository a first-class Kubernetes resource, and it splits a backup into three resources: the recipe, its invocation, and its schedule. That is what lets a backup be triggered by cron, by kubectl create, by Argo Events, or by a Helm hook. A kopia snapshot's lifecycle is tied to its Snapshot object through a finalizer and a deletionPolicy.

The whole CRD surface is modeled as Rust enums, so invalid states cannot be expressed and reconcilers handle every variant at compile time. For the high-level mental model start with Concepts.

Alpha

API group kopiur.home-operations.com, version v1alpha1. The CRD surface may still change between releases.

The 9 CRDs (kopiur.home-operations.com/v1alpha1)

CRD	Scope	Layer	Purpose
Repository	Namespaced	Storage	A kopia repository owned by one namespace: backend, encryption, credentials.
ClusterRepository	Cluster	Storage	A shared repository for platform teams, gated by allowedNamespaces.
SnapshotPolicy	Namespaced	Recipe	What to back up: PVC sources, identity, retention, policy, hooks, into one repository or a 1–8 multi-repository fan-out. Idempotent.
Snapshot	Namespaced	Invocation + Catalog	One kopia snapshot as a Kubernetes object. The universal trigger entry point.
SnapshotSchedule	Namespaced	Cron	When it runs: cron + jitter + timezone; creates Snapshot CRs.
Restore	Namespaced	Operation	Restore a snapshot to a PVC, or act as a passive volume-populator source.
Maintenance	Namespaced	Lifecycle	Schedules kopia maintenance quick + full with an ownership lease.
RepositoryReplication	Namespaced	Durability	Mirror a repository's blobs to a second backend on a schedule (the "2" in 3-2-1).
SnapshotReplication	Namespaced	Durability	Copy selected snapshots from one repository into another on a schedule (kopia snapshot migrate).

Where to next

• How Kopia works covers content-addressable dedup, snapshots, the username@hostname:path identity model, encryption and maintenance, plus why one shared repository is the recommended layout.
• Why Kopiur is designed this way covers the recipe, invocation and schedule split, repository-as-resource, the type-safety argument, and tying a snapshot's lifecycle to its object.
• Getting started is the end-to-end walkthrough: install, first backup, and a verified restore in about 15 minutes.
• Scenarios are problem-driven, end-to-end walkthroughs: protect a database, recover deleted data, disaster recovery, migration, adopting an existing repository, verification drills.
• Installation covers prerequisites, install modes, and the CRD-lifecycle caveat.
• Repositories & backends points Kopiur at S3, Azure, GCS, B2, a NAS, or rclone.
• Backups & schedules and Restores cover the recipe, invocation and schedule model, and reading data back.
• Troubleshooting is for when something does not go green.
• GitOps (Flux / Argo) covers kstatus health, kubectl wait, managed-by and ownerRefs, and drift-free applies.
• Field reference lists every field of all 9 CRDs, with type, default and immutability.
• API reference (rustdoc) is the generated Rust API docs for every crate in the workspace.
• API conventions and Observability are developer notes.