Storage Management

Zyvor Fabric supports multiple storage backends and provides a unified API for volume management, snapshots, and cloning.

---

Storage Backends

Backend	Description	Best For
Local	Default filesystem storage	Single-node, development
NFS	Network filesystem	Shared storage, multi-node
LVM	Logical Volume Manager	Flexible local storage
LVM-thin	Thin-provisioned LVM	Overcommitted storage
ZFS	ZFS datasets with replication	Snapshots, data integrity
Ceph/RBD	Distributed block storage	HA, multi-node, production

Configuration

# /etc/zyvor-fabricd/zyvor-fabricd.toml

# Local (default)
[storage]
backend = "local"
path = "/var/lib/zyvor-fabricd/volumes"

# NFS
[storage]
backend = "nfs"
server = "nfs.example.com"
export = "/exports/zyvor-fabricd"
mount_point = "/mnt/zyvor-fabricd"

# Ceph/RBD
[storage]
backend = "rbd"
pool = "Zyvor Fabric"
monitors = ["mon1.example.com", "mon2.example.com"]

See NFS_STORAGE_GUIDE.md for detailed NFS setup.

---

Volumes

Create

curl -X POST http://localhost:9095/api/volumes \
  -H "Content-Type: application/json" \
  -d '{"name": "data-volume", "size_gb": 100, "format": "qcow2"}'

List

curl http://localhost:9095/api/volumes

Resize

curl -X POST http://localhost:9095/api/volumes/data-volume/resize \
  -H "Content-Type: application/json" \
  -d '{"size_gb": 200}'

Clone

curl -X POST http://localhost:9095/api/volumes/data-volume/clone \
  -H "Content-Type: application/json" \
  -d '{"name": "data-volume-copy"}'

---

Snapshots

Create

curl -X POST http://localhost:9095/api/volumes/data-volume/snapshots \
  -H "Content-Type: application/json" \
  -d '{"name": "backup-2026-02-18"}'

List

curl http://localhost:9095/api/volumes/data-volume/snapshots

Restore

curl -X POST http://localhost:9095/api/volumes/data-volume/restore \
  -H "Content-Type: application/json" \
  -d '{"snapshot_id": "abc123"}'

---

Ceph/RBD

CLI (fabricctl)

# Create a Ceph storage pool
fabricctl ceph create my-pool \
  --monitors=10.0.0.1,10.0.0.2,10.0.0.3 \
  --pool=rbd \
  --user=admin \
  --keyring=/etc/ceph/ceph.client.admin.keyring

# Health and stats
fabricctl ceph health my-pool
fabricctl ceph stats my-pool

# RBD image management
fabricctl ceph images my-pool
fabricctl ceph create-image my-pool vm-disk-01 --size=10240
fabricctl ceph delete-image my-pool vm-disk-01

# Export as YAML
fabricctl ceph pools -o yaml

API

# Create pool
curl -X POST http://localhost:9095/api/storage/pools/ceph \
  -H "Content-Type: application/json" \
  -d '{
    "name": "my-pool",
    "monitors": ["10.0.0.1", "10.0.0.2", "10.0.0.3"],
    "pool_name": "rbd",
    "user": "admin",
    "keyring": "/etc/ceph/ceph.client.admin.keyring",
    "auto_start": true
  }'

# Health, stats, images
curl http://localhost:9095/api/storage/pools/my-pool/health
curl http://localhost:9095/api/storage/pools/my-pool/stats
curl http://localhost:9095/api/storage/pools/my-pool/images

# Create RBD image
curl -X POST http://localhost:9095/api/storage/pools/my-pool/images \
  -H "Content-Type: application/json" \
  -d '{"name": "vm-disk-01", "size_mb": 10240}'

Web UI

Navigate to Storage Pools > Create Pool > select Ceph. Configure monitor addresses, pool name, user, and keyring. Ceph pools display health status (Ok/Warn/Error) in the pool list.

---

Disk Formats

Format	Snapshots	Performance	Thin Provisioning	Compatibility
qcow2	Yes	Good	Yes	QEMU (default)
raw	No	Best	No	Universal
vmdk	No	Good	Varies	VMware
vdi	No	Good	Varies	VirtualBox

---

Best Practices

1. Use qcow2 for production -- supports snapshots, compression, and thin provisioning
2. Use raw for performance-critical workloads where snapshot support is not needed
3. Take snapshots before major changes -- they are fast and cheap with qcow2
4. Monitor disk usage -- set up alerts to prevent running out of space
5. Use thin provisioning to overcommit storage where workloads permit
6. Back up important volumes regularly and test restores