Storage that keeps working when a server fails.

Haify turns the disks in your own Linux servers into shared storage for files, virtual machines, containers and databases. Lose a server and your applications keep their data and carry on.

Get started Source on GitHub
files VMs containers databases one address server 1 in sync server 2 in sync server 3 in sync

What people use it for

share server 1 server 2

A shared folder that never goes down

Team files, media libraries, build artifacts. Clients keep the same address, and a failed server no longer means an outage.

Set up a share
host 1 host 2 same disk on both VM

Virtual machines that outlive their host

VM disks kept on two servers, so you can move a running VM between them and restart it on the other one if its host dies. On Proxmox VE, XCP‑ng, plain KVM or OpenStack.

Proxmox VE XCP-ng KVM OpenStack
node 1 node 2 node 3 app app

Databases and apps in containers

Persistent volumes that outlive the node they started on. A pod or container rescheduled elsewhere comes back with its data.

Kubernetes Helm chart Docker
app server 1 server 2 DB

A database that fails over on its own

PostgreSQL, MySQL or any service gets its data, its address and itself moved to a healthy server, without changing the application.

Make a service highly available
your cluster backups second site

Backups and a second site

Scheduled backups to S3, SMB or WebDAV, and a live copy in another location. Recover even if the whole cluster is lost.

Protect your data
storage array server server server

Leaving a storage array behind

Get the availability you paid an array for from ordinary servers and disks, with block shares your existing hosts already understand.

Plan a cluster

Why Haify

No single point of failure
Every write is on more than one server, and services move on their own when one fails.
Your hardware
Ordinary Linux servers and their local disks. No storage array, no special network.
Simple to run
Manage it from a web UI, the command line or an AI assistant.
Problems found early
Alerts in your chat when something degrades, and a daily check that says how to fix it.
Open source
Apache 2.0, free to use, change and run in production.

From disks to a shared folder

Four commands, once Haify is installed.

  1. Add your servers.

    haify node register --name node1 --address 192.0.2.11 haify node register --name node2 --address 192.0.2.12

  2. Give Haify a disk on each one.

    haify pool create --name pool0 --type lvm-thin --nodes node1,node2 --devices /dev/sdb

  3. Create a 10 GiB volume kept on both.

    haify resource create --name data --port 7001 --size 10G --nodes node1,node2 --pool pool0

  4. Share it over NFS.

    haify gateway nfs create --resource data --service-ip 192.0.2.200/24 --export-path /data

Documentation