VPS Snaps

Linode backups and snapshots on Akamai Cloud: enable, snapshot and restore

On Akamai Cloud (formerly Linode), snapshots are part of the paid Backups service. With it on, each Linode gets three automatic backups (daily, weekly and one up to two weeks old) plus one manual snapshot slot. Take the snapshot from the Linode's Backups tab or with linode-cli linodes snapshot 123 --label before-upgrade, and restore it to the same Linode or another one in the same region. The service only backs up unencrypted ext3 and ext4 disks.

7 min readUpdated Checked against official documentation

How the Backups service works

Each backup is a full, file-based snapshot of the Linode's disks, taken while the Linode keeps running, in a two-hour window you choose. The service keeps up to four:

SlotAgeHow it is filled
DailyLess than 24 hours oldAutomatically, in your chosen window
WeeklyLess than 7 days oldThe daily backup from the weekday you choose
Biweekly8 to 14 days oldThe previous weekly backup moves here
Manual snapshotWhenever you took itBy you; replaced by your next snapshot

Backups are stored on separate hardware in the same data center as the Linode. The service runs in all core compute regions but not in distributed compute regions. Because backups are file-based rather than block images, they come with filesystem rules, listed under Limits below.

Enable backups

  1. Log in to Cloud Manager and open Linodes.
  2. Select the Linode and open its Backups tab.
  3. Click Enable Backups and confirm.

To enroll every new Linode automatically, turn on the Backup Auto Enrollment switch in your account settings. It doesn't enroll Linodes that already exist.

From the CLI, find the Linode's ID, then enable the service. The CLI reads a token from linode-cli configure or the LINODE_CLI_TOKEN environment variable.

Terminal
linode-cli linodes list --format id,label,region
Terminal
linode-cli linodes backups-enable 123

To set the schedule, open the Backups tab, choose a Time of Day window and a Day of Week for the weekly slot under Settings, and click Save Schedule.

Take a manual snapshot

  1. In Cloud Manager, open Linodes and select the Linode.
  2. Go to the Backups tab.
  3. Under Manual Snapshot, enter a label and click Take Snapshot.
  4. Confirm in the dialog.
Terminal
linode-cli linodes snapshot 123 --label before-upgrade
  • 123 is the Linode ID.
  • --label names the snapshot, 1 to 255 characters. The API requires it.

There is only one manual slot. Taking a new snapshot deletes the previous one, so the manual snapshot is the latest checkpoint, not a history.

A snapshot takes several minutes or longer, depending on how much data the Linode holds. Other Cloud Manager jobs for that Linode wait until it finishes.

List backups and check consistency

Terminal
linode-cli linodes backups-list 123

The response lists the automatic backups and, under snapshot, the current manual snapshot and any snapshot in progress. Each entry carries the backup ID a restore needs. To see one in detail:

Terminal
linode-cli linodes backup-view 123 456789

Backups run while the Linode is up, and Akamai notes that a database caught mid-transaction can end up in an unclean state. Dump the database before the backup window or before a manual snapshot, with pg_dump, mysqldump or mongodump, so a consistent copy sits in a file the backup picks up.

Restore to the same or another Linode

  1. Open the source Linode's Backups tab.
  2. Find the backup or snapshot and choose Restore to Existing Linode from its menu.
  3. Pick the target Linode. It can be the same Linode or another one in the same region.
  4. Check the Space Required column. The target needs that much unallocated disk space, unless you choose to overwrite it.
  5. Confirm and wait for the restore to finish.
Terminal
linode-cli linodes backup-restore 123 456789 --linode_id 234 --overwrite true
  • 123 is the source Linode and 456789 the backup or snapshot ID.
  • --linode_id is the target Linode. Use the source's own ID to restore it in place.
  • --overwrite true deletes all disks and configuration profiles on the target first. With false, the restore fails if the target lacks the space.

A restored disk gets the same UUID as the original. If both disks are attached to the same Linode at once, only one mounts, picked at random, and the Linode may fail to boot. To work with both copies, restore to a separate Linode.

Restore to a new Linode

  1. On the source Linode's Backups tab, choose Deploy New Linode from the backup's menu.
  2. The create page opens with Create from Backup selected and the source filled in. Pick a plan, enter a label and click Create.
Terminal
linode-cli linodes create --backup_id 456789 --type g6-standard-2 --region us-east --label web-01-restore
  • --backup_id is the backup or snapshot to use. It must have completed successfully.
  • --region must match the backup's region. Backups can't be restored across regions, but you can migrate the new Linode afterwards.
  • --type is the plan, which must be large enough for the backup.
  • The new Linode gets the same disks, configuration and root password as the one that was backed up.

Restores take from a few minutes to several hours, depending on size. A restore brings back whole disks. To recover one file, restore to a new Linode and copy the file across.

Automate snapshots

The automatic slots already give you daily, weekly and two-week-old restore points. A scheduled manual snapshot adds a fourth at a time you choose, such as just before a nightly deploy. Each run replaces the previous snapshot.

/usr/local/bin/linode-snapshot.sh
#!/usr/bin/env bash
set -euo pipefail

LINODE_ID=123

linode-cli linodes snapshot "$LINODE_ID" --label "auto-$(date -u +%Y%m%d-%H%M)"
/etc/cron.d/linode-snapshot
45 1 * * * root /usr/local/bin/linode-snapshot.sh >> /var/log/linode-snapshot.log 2>&1

Set LINODE_CLI_TOKEN in the script, or run linode-cli configure as root. If linode-cli was installed with pip or pipx, cron may not find it, so use the full path from command -v linode-cli.

If you need more than one manual restore point, or history older than 14 days, the Backups service can't hold it. That history has to live somewhere else.

What it costs

The Backups service is a paid add-on priced per Linode plan, with hourly and monthly rates, and some regions are priced differently. As of October 2026, Akamai's documentation points to its pricing page for the rates rather than listing them. The API's List types endpoint also returns each plan's backup price, including region-specific prices.

The manual snapshot is part of the Backups service, not a separately billed product. Cancelling the service deletes every backup, including the manual snapshot.

Limits

  • Disks must hold unencrypted ext3 or ext4 filesystems the service can mount. Disks encrypted inside the operating system, or with multiple partitions, can't be backed up.
  • Very large numbers of files slow backups down and can make them fail.
  • ACLs and extended attributes aren't tracked.
  • A file changed without changing its size or modification time isn't seen as changed in the next backup.
  • Block Storage volumes are never included, even when attached.
  • Backups can't be restored across regions and can't be downloaded directly.
  • Backups aren't encrypted, even when taken from a Linode that uses disk encryption. Data is encrypted again on restore if encryption is enabled.
  • Deleting a Linode deletes its backups.

What backups and snapshots do not protect against

Every backup slot sits in the same Akamai account and the same data center as the Linode.

  • Account loss. A suspended or closed account takes its Linodes and their backups with it.
  • Deleting the Linode or cancelling backups. Either one removes every copy at once.
  • Problems you find late. The oldest automatic backup is at most 14 days old, so data that went bad three weeks ago may have no clean copy.
  • Data the service skips. Block Storage, encrypted or multi-partition disks, ACLs and extended attributes aren't backed up.

Akamai itself advises backing up your data independently before you cancel the service. Do it regardless: send database dumps and file archives to storage at another provider, or to a machine you control. That is the off-site leg of the 3-2-1 rule. Restore from it now and then to prove it works.

A snapshot is the whole disk, but not a deleted account or a change you notice too late. For what else to keep, see what to back up on a Linux server, and for how often, RPO and RTO explained.

Frequently asked questions

How many manual snapshots can a Linode have?
One. Taking a new manual snapshot deletes the previous one.
Can I download a Linode backup?
Not directly. Restore it to a Linode, then copy files off with rsync, SCP or a similar tool.
Does the Backups service include Block Storage volumes?
No, even when a volume is attached to the Linode.
Can I restore a backup to a Linode in another region?
No. Restore to a Linode in the same region, then migrate it if you need it elsewhere.
What happens to backups if I delete the Linode?
They are deleted with it. Copy anything you need to keep somewhere else first.

How this was checked

Commands, limits and prices were checked against these official pages, on October 3, 2026: