Backup & Snapshots

OpenIDCS offers two data protection mechanisms for virtual machines — snapshots and full backups — so you can recover quickly from accidental changes, system corruption or migration.

Comparison

Aspect	Snapshot	Backup
Storage	Shares the original VM storage pool	Independent backup directory (can be mounted remote storage)
Speed	Seconds (copy-on-write)	Minutes to hours, depending on VM size
Space usage	Incremental; very small at first	Full copy, comparable to the original VM
VM deletion	Snapshots become unusable	Backup remains independent; VM can be rebuilt
Primary use	Pre-change rollback point, ad-hoc test	Long-term archive, DR, cross-host migration
Suggested cadence	Manual before critical operations	Scheduled daily / weekly

Best Practice

Snapshots are for short-term protection (e.g. before a system update). Backups are for long-term protection (e.g. archiving business data). Use them together — never replace backup with snapshot.

Platform Matrix

Platform	Snapshot	Backup	Restore	List backups	Delete backup
VMware Workstation	✅	✅	✅	✅	✅
VMware vSphere ESXi	✅	✅	✅	✅	✅
LXC / LXD	✅	✅	✅	✅	✅
Docker / Podman	✅ (commit)	✅	✅	✅	✅
Proxmox VE	✅	✅	✅	⚠️ node-local	⚠️ node-local
Windows Hyper-V	✅	✅	✅	✅	✅
Qingzhou Cloud	✅	✅	✅	✅	✅

Snapshot Management

Create

1. VM Management → pick the VM → Detail.
2. Switch to the Snapshots tab.
3. Click Create Snapshot and fill in:
   • Name: e.g. pre-upgrade-2026-04-24
   • Description: why
   • Include memory: capture runtime memory (requires the VM to be running)
4. Click Confirm.

Note

• Capturing memory briefly freezes the VM for a few seconds.
• VMware/ESXi/Hyper-V support multi-layer snapshot trees; Docker commit produces flat image layers.

Restore

1. Select the target snapshot in the list.
2. Click Restore to this snapshot.
3. The system powers off the VM → rolls back disks → powers on again.

Data Loss Risk

Restoring a snapshot discards all disk changes made after it was created. Back up anything important first.

Delete

• Delete single: reclaim the incremental layer.
• Consolidate all: merge every snapshot into the current disk (VMware/ESXi only).

TIP

In production, keep no more than 3 snapshots per VM — beyond that, disk I/O degrades significantly.

Naming Convention (suggested)

<action>-<date>-<note>
# Example
pre-upgrade-20260424-kernel
before-patch-20260424-cve
daily-auto-20260424

Full Backup

Manual Backup

1. VM detail → Backup tab.
2. Click Create Backup.
3. Fill in:

Parameter	Description	Recommendation
Backup name	Filename prefix	Use a meaningful name
Description	Free-form note	Record business state
Compression	0 (none) to 9 (max)	6 balances speed/size
Include snapshots	Bundle existing snapshots	Recommended for prod
Shutdown before backup	Power off first	Recommended for DB-like VMs

4. Click Start Backup; track progress in the Task Center.

Backup List

The Backup tab shows:

Field	Meaning
Backup ID	Unique identifier
Created	Start time
Size	Size after compression
Status	Completed / In progress / Failed
Actions	Restore / Download / Delete

Restore

1. Pick a backup and click Restore.
2. Pick the mode:
   • Overwrite current VM: rollback in place.
   • Restore as new VM: create a new VM with new name/IP (good for validation).
3. Confirm and track progress in the Task Center.

Overwrite destroys current data

Overwrite restore wipes the current VM's disks. Double-check before confirming.

Download

Click Download to fetch the archive locally, typically:

• .tar.gz (LXD / Docker)
• .ova / .vmdk (VMware / ESXi)
• .vhdx (Hyper-V)
• .vma.zst (Proxmox)

Scheduled Backup Strategy

Create a Schedule

1. System Settings → Scheduled Tasks.
2. Click New Task → type Backup.
3. Configure:

name: daily-backup-prod
targets:
  - tag: env=production
  - or a list of VM IDs
cron: "0 2 * * *"    # every day 02:00
compression: 6
retention:
  - keep 7 daily backups
  - keep 4 weekly backups
  - keep 12 monthly backups
notify: email on success or failure

4. Save — the task enters the scheduler queue.

3-2-1 Backup Rule (recommended)

3 copies of data (1 production + 2 backups)
2 different storage media (local disk + object storage)
1 offsite copy (remote DC / cloud)

Offsite Backup

Mount remote targets under System Settings → Storage:

Type	Fields
NFS	Server, export path, mount point
SMB/CIFS	Server, share name, user/password
S3-compatible	Endpoint, bucket, access/secret key
SFTP	Host, port, user, private key

Once mounted, tick Sync to remote storage in the scheduled task.

Cross-host Migration

Backups enable migration across hypervisors or hosts:

graph LR
    A[Source host A] -->|create backup| B[Backup file]
    B -->|download/sync| C[Target host B]
    C -->|restore as new VM| D[Target VM]

Steps

1. Create a full backup on the source (power off for consistency).
2. Download the file to the controller.
3. Upload to the target host → Restore as new VM.
4. Adjust network settings (IP / bridge) after restore.

TIP

Same-platform restore (VMware → VMware) is direct. Cross-platform (e.g. VMware → KVM) requires converting the disk format first, e.g. qemu-img convert -f vmdk -O qcow2 src.vmdk dst.qcow2.

Backup Validation

Regularly verify that backups really restore. Suggested monthly exercise:

1. Pick 1–2 representative VMs from production backups.
2. Restore them with Restore as new VM on an isolated host.
3. Boot the VM and verify:
   • OS boots cleanly
   • Critical services respond
   • Data integrity (DB row counts, file MD5)
4. Delete the test VM when done.

API Examples

Create a backup

curl -X POST http://localhost:1880/api/vms/{vm_id}/backup \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "manual-20260424",
    "description": "release v2.1 upgrade",
    "compress": 6,
    "include_snapshots": true
  }'

List backups

curl http://localhost:1880/api/vms/{vm_id}/backups \
  -H "Authorization: Bearer YOUR_TOKEN"

Restore a backup

curl -X POST http://localhost:1880/api/vms/{vm_id}/restore \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"backup_id": "bak_20260424_020000", "mode": "overwrite"}'

Troubleshooting

Symptom	Likely Cause	Fix
Backup task stalls	Disk I/O contention	Stagger schedules, cap parallelism
Backup file is unexpectedly large	Large zero regions inside the disk	Enable sparse backup, run fstrim regularly
Restored VM will not boot	Disk format incompatible	Convert with qemu-img
Snapshot count capped	Platform limit (e.g. VMware 32 levels)	Consolidate old snapshots
Remote sync fails	Network / cert issue	Check the mount and task logs

Related Docs

• 📋 Virtual Machine Management
• 📊 Monitoring & Alerts
• 📝 Logs & Audit
• ⚙️ Server Setup