VM and disk backup and recovery
Snapshots prove a copy exists, not that the instance comes back. Eon backs up VMs and their disks on AWS, Google Cloud and Azure as whole instances, and restores the machine or one file from it.
What Eon backs up
Instances and their attached disks, discovered across every account, project or subscription you connect.
AWS
Amazon EC2 and Amazon EBS. Instance configuration, attached volumes and AMI-level capture. Restore a whole instance, a single volume out of an instance backup, or one file. Instant Restore boots the instance before every volume has finished hydrating.
Boot first, hydrate after
The instance serves traffic before the disks are full.
Instant Restore boots the instance right away and hydrates the rest in the background, so time to a usable machine no longer grows with disk size.
Bulk restore for fleets.
Recover many instances in one operation, on any cloud, with progress and an estimated finish time you can share.
One file, not the whole machine.
The backup is indexed, so you search for the file or directory you need and restore only that, with no instance to stand up.
A snapshot is not a recovery plan
A 10 TB instance on four volumes means matching four snapshots from the same moment, rebuilding and attaching each volume, then waiting for every block to hydrate. Eon snapshots all the volumes as one group and restores the machine with its device mapping, private IP and MAC address intact.
Native snapshots
Eon
What gets captured
Volumes, one at a time. Reassembling an instance is on you.
The instance, its disks and their relationships, captured together.
Time to a usable machine
Degraded until every block has hydrated, so it scales with disk size.
Instant Restore boots the instance and serves traffic while the rest catches up.
Getting one file back
Restore the volume, attach it, mount it, then go looking.
Search the backup and pull back the file. The instance stays untouched.
Databases on those disks
Crash-consistent, so the engine starts in recovery.
Handled as databases, with restore down to the rows a query matches.
Proof the backup restores
The job reports success. Whether the volume actually reads back is not checked.
Every backup is turned into a volume, mounted and scanned. It passes only if the filesystem reads cleanly.
Knowing what is protected
Tags and configuration you maintain by hand.
New instances discovered on their own. Drift shows up as a finding.
Ransomware and rogue AI agents
In both cases something with valid credentials is destroying data, and backups are an early target. Eon keeps recovery data in an isolated vault those credentials cannot reach, and scans every backup so you know which copy is clean.
Credentials from production do not open your backups
Instance backups land in an isolated, immutable vault, kept apart from the account they came from. Access is time-bound, so a key stolen out of production does not reach your recovery data, and nothing in there can be quietly overwritten or deleted.
You know which copy is clean before you restore
Every backup is scanned as it lands, for entropy shifts, mass deletion, suspicious file changes and ransomware notes. You pick a recovery point you know is good, rather than booting one and finding out.
Recover a fleet, not one machine at a time
When an attack crosses many instances, bulk restore brings them back in one operation, with progress and an estimated finish time you can share while it runs.
What VM backup costs
Snapshots outlive their instances, every Region adds a copy, and retention only ever grows. Eon stores changes instead of copies, deduplicates across the estate, and flags orphaned snapshots and volumes before they pile up on the bill.
Databases on those instances stay queryable
A supported self-managed database on the instance is backed up as a database, in Apache Iceberg and Parquet tables. Query them with analytics engines like Athena, BigQuery or Microsoft Fabric, or ask in plain English, with no restore. Everything else on the instance restores at file level.
More on VM and disk backup
FAQs
Usually nothing. Eon connects through a role, service account or service principal and backs up through the cloud APIs, with no appliances or inbound ports. For a lower RPO on specific instances, you can turn on an optional lightweight component for those workloads only.
The instance boots and starts serving before hydration finishes, so time to a usable machine does not scale with disk size the way a conventional restore does.
Yes to both. Restore one volume out of an instance backup, or search the backup and pull back individual files without restoring the instance at all.
Yes, and to a different project or subscription. This is also how recovery drills are run, since you can restore somewhere isolated and confirm the machine comes up without touching production.
Crash-consistent snapshots are a real risk for databases, because the snapshot does not coordinate with the engine. Eon handles databases on instance volumes as databases, with restore down to the rows a query matches, rather than treating them as one more file on a volume.
Yes. On Azure that includes disks encrypted with Azure Disk Encryption, on both the backup and the restore path, which Azure Disk Backup does not cover.
See it against your own instances
Connect a read-only role to see every instance and disk you have, what is protected, and which snapshots outlived their source. Bring a recovery you have had to do and we will show you how it runs on Eon.