Adding Clusters

Deploy a New Cluster

The Deploy a new cluster flow can prepare a local single-node deployment on this computer, or install Buckit on fresh Linux hosts you can reach over SSH and form them into a new cluster.

If Buckit Manager runs on macOS or Windows, it first asks where Buckit should be deployed:

• This computer only — download and configure Buckit to run on this computer only as a single-node cluster.

• Remote servers — deploy Buckit as a managed cluster on one or more Linux servers over SSH.

If Buckit Manager runs on Linux, it opens the remote servers wizard directly.

The remote servers wizard guides you through these steps:

1. Basics — name the cluster and choose the Buckit version, root credentials, S3 API and console ports, region, and optional TLS certificate and key.

2. Nodes — enter the target hostnames and the SSH credentials Buckit Manager uses to reach them. You can override SSH settings per host.

3. Discover — Buckit Manager connects to each host over SSH and collects the operating system, CPU, memory, and mounted drives.

4. Topology — choose which data drives to use and confirm the erasure-coding set size and parity for the cluster layout.

5. Preflight — Buckit Manager validates each host. Blocking errors must be resolved before continuing; advisory warnings do not stop the deploy.

6. Review — review the full deployment plan.

7. Deploy — Buckit Manager downloads the package, installs Buckit, writes configuration, starts the service, and waits for health on each host, showing per-node progress.

8. Done — a summary of the new cluster.

For a step-by-step walkthrough of this wizard, see Guided Deployment using Buckit Manager.

Import an Existing Cluster

The Import existing cluster flow registers a running Buckit or MinIO deployment so you can monitor and operate it from Buckit Manager.

Provide the cluster’s S3 endpoint and root credentials:

• Cluster URL — the S3 endpoint of any node in the cluster, or the load balancer if you have one.

• Access key and Secret key — the root credentials the server was started with.

Buckit Manager contacts the admin API and discovers every node, drive, and pool. While discovery runs, the interface streams progress. When discovery finishes, it shows a summary — engine (Buckit or MinIO), version, node, pool, and drive counts, parity, and capacity — and lets you name the cluster before saving. After you save, Buckit Manager opens the cluster detail page.

Migrate MinIO to Buckit

Clusters imported as MinIO Community Edition can be upgraded in place to the latest Buckit — same disks, same data. Start the migration from the Migrate to Buckit button on the detail page of an imported MinIO cluster.

The migration wizard guides you through these steps:

1. Overview — choose the target Buckit version and review the plan.

2. SSH credentials — provide the SSH credentials and per-host ports Buckit Manager uses to run the cutover. These are saved to the cluster so later operations can reuse them.

3. Preflight Check — Buckit Manager takes a snapshot of the MinIO deployment’s configuration (buckets, users, groups, policies, service accounts, lifecycle rules, notifications, and replication targets) and runs preflight validation. Blocking failures must be resolved before continuing.

4. Migrate — Buckit Manager performs the cutover node by node: stopping MinIO, installing the Buckit package, switching the systemd unit, and waiting for health. It then verifies that the migrated cluster’s buckets, objects, and identity configuration match the snapshot, and rolls back on failure.

For a complete migration guide, including prerequisites and rollback behavior, see Guided Migration using Buckit Manager.