Watch: a Ceph cluster
Bootstrap, OSDs, pools, CephFS and RBD, and day-two operations. The written steps are below.
Ceph From the Dashboard
Ceph is the gold standard for distributed storage — and famously unfriendly to set up. WolfStack puts the whole lifecycle in the dashboard: install the packages, bootstrap a cluster, add disks, carve out pools, and serve CephFS filesystems or RBD block images, all without memorising ceph CLI incantations.
What You Can Do
| Task | From the dashboard |
|---|---|
| Install & bootstrap | WolfStack checks which Ceph components are installed per node, installs what's missing (with your confirmation — nothing installs silently), and bootstraps a new cluster. |
| OSDs (disks) | Lists eligible devices per node, creates OSDs on them, removes OSDs, and handles day-2 actions: mark in/out, reweight. |
| Pools | Create, tune (replica size and options) and delete pools. |
| CephFS | Create and remove CephFS filesystems — mountable from any node or shared onward via Shares. |
| RBD images | List, create and delete RADOS block images per pool, for use by Ceph clients. WolfStack's own VMs don't attach RBD images directly — to keep VM disks on Ceph, mount a CephFS filesystem and choose it as the VM's storage. |
| Monitors & managers | Add and manage extra mon and mgr daemons for high availability, and start/stop/restart Ceph daemons per node. |
| Fix & repair | Day-2 recovery without the CLI: set and clear cluster flags (noout and friends), scrub and repair placement groups, run the balancer, and repair OSDs. |
| Health & recovery | The cluster status view shows overall health, monitor quorum, OSD states, MDS (CephFS metadata) status and live recovery/backfill progress, with capacity at a glance. |
Bootstrapping a Ceph cluster — step by step
- Open a node in the sidebar and choose Ceph. WolfStack checks which Ceph components each node has and offers to install anything missing — confirm the install (nothing installs silently).
- Click Bootstrap cluster on your first node to create the initial monitor and manager.
- Add OSDs: WolfStack lists the eligible disks on each node — select the ones to dedicate to Ceph and create OSDs on them. Three nodes with disks is the practical minimum for self-healing replication.
- Add extra monitors and managers on the other nodes for high availability.
- Create a pool (set its replica size), then a CephFS filesystem or RBD image on top — CephFS mounts on any node (and can be re-shared via Shares) and can hold VM disks as a VM storage location; RBD images are for other Ceph clients — WolfStack VMs don't use them directly.
Day-2 work — setting noout before maintenance, scrubbing and repairing placement groups, reweighting or replacing an OSD — is all in the same view, so you rarely need the Ceph CLI.
When to Use Ceph vs the Alternatives
- Ceph — three or more nodes, lots of disks, and you want self-healing replicated storage with block, file and object access. The heavyweight champion; needs RAM and patience.
- WolfDisk — our lighter-weight replicated filesystem. Far simpler to run for small clusters that just want shared, replicated directories.
- Storage Array — single-node parity arrays (mdadm/NoNRAID) when the redundancy you need is within one box.
Try this on your own servers
WolfStack is free for personal use and installs in about two minutes — one command, no agents, works on any Linux server or Raspberry Pi.
Running it at work? Commercial licences start small and fund development.