Cloud Server Backups – NABHCLOUD

Guides

Server Backup keeps rolling restore points for your cloud server, so a crashed disk, a compromised site or an accidentally removed file means a quick rollback rather than rebuilding from scratch. Every plan includes at least 7 daily plus 7 weekly restore points, along with Vault Protect, and current pricing is shown on this page.

Explore our VPS plans — hourly billing, GST invoices, and servers ready in under a minute.

What's included

Each protected server comes with at least 7 daily plus 7 weekly restore points as standard. You can increase these limits for an additional monthly charge per extra point (current pricing is shown on this page):

• Daily restore points: up to a maximum of 30.
• Weekly restore points: up to a maximum of 12.
• Monthly restore points: up to a maximum of 12.

A weekly or monthly restore point is simply that day's daily backup held for longer, so storage is not duplicated.

Server Backup and Vault Protect become available once an account has at least one completed payment. Backups run at a time of day you select (IST), and charges are deducted hourly from your wallet, in the same way as the server itself. You may also trigger one manual backup per day, which stays available for 24 hours in addition to the scheduled points.

You can switch this feature on at order time, either from the Build page or from the order page itself. On an account that hasn't yet made a payment, it activates automatically as soon as the first payment goes through, and the setting stays held for up to 30 days before that happens.

How backups are stored

Restore points use incremental storage: weekly and monthly restore points land as complete copies, while daily points capture only the portions of data that have changed since the previous backup, rather than storing everything again. Everything gets compressed and locked with an encryption key that's unique to that particular server, then held in an S3-style backup bucket inside your account. Storage expands automatically in 100 GB increments whenever backups need more space — you'll get an email explaining why — and contracts again once older points expire.

Restoring a backup

In-place restore rewinds the server to whichever point you select; before it happens, we keep a snapshot of the disk's current state for 24 hours, so the rewind itself can be reversed. Building a server from a chosen restore point instead creates a brand-new server with its own IP address, separate from the original — handy for testing without risk to the live system. You can also download the disk image directly. None of these restore options carry an extra charge.

The Browse files option opens a restore point in read-only mode inside the control panel. You can look through its folders or search by filename, tick the files you want, then package them into a zip you can grab within the next 24 hours. Your live server stays untouched throughout.

Backups that respect running applications

When the QEMU guest agent runs inside the server — installed via apt install qemu-guest-agent, dnf install qemu-guest-agent, or, for Windows servers, the virtio-win guest tools that depend on VSS — the backup process holds off disk writes for an instant, giving databases and open files time to reach a clean, consistent state before the snapshot is taken. Without the agent, a restore point is crash-consistent instead — equivalent to the state right after a sudden power loss, something journaling filesystems and databases are designed to recover from automatically when they boot. The control panel shows which type each restore point is.

Checking that restores actually work

Each month, we boot a selection of backups on an isolated network to confirm they restore correctly; any server that passes is marked Verified restorable. You can also run this check yourself once a month at no extra cost — we boot the latest restore point alongside your live server without disturbing it. You then get console access for an hour.

Managing backups via API and MCP

Add protection: backup to a POST /api/v1/services call (or use the order_servers MCP tool) to order it together with the server. Use PUT /api/v1/services/<id>/backup to turn the feature on or change its schedule, GET /api/v1/services/<id>/backup/points to list restore points, and POST /api/v1/services/<id>/backup/points/<point>/restore to restore one of them. The equivalent MCP tools are set_server_backup, list_restore_points, backup_server_now and restore_backup_point.

Common questions about backups