Documentation
Documentation for cetrafs, the Linux shared-disk cluster filesystem: tutorials, how-to guides, design concepts, and reference for the tools and on-disk format.
The documentation is split into four kinds, according to what you need at the moment.
I want to learn
Guided, hands-on lessons that end with something working. Start with a single-node volume on a loop device; it needs no special hardware.
How-to guidesI have a task
Step-by-step recipes for installing, configuring, monitoring, repairing and recovering a cetrafs deployment.
ConceptsI want to understand
How cetrafs coordinates nodes through the storage, why it locks whole files, what fencing does and doesn't cover, and how it compares to VMFS and GFS2.
ReferenceI need a fact
Tool flags and exit codes, mount options, every procfs field and module parameter, hard limits, and the on-disk format.
Where to start
- Evaluating cetrafs? Read What is cetrafs?, then Project status and Limits and unsupported operations. Those three pages will tell you whether it fits your workload. Quick answers are in the FAQ.
- Want to try it? Follow Your first cetrafs volume. It runs on any supported Linux host using a loop device.
- Deploying on Proxmox VE? See how it compares in Shared storage for Proxmox VE without Ceph, then work through Build a Proxmox VE shared datastore, then the VM disk settings guide.
- Something went wrong? See Troubleshooting and Recover from a node failure.