A Successful Home Assistant Backup: Will It Restore on Another Machine?

A successful Home Assistant backup does not prove full recovery. Check included data, encryption keys, external databases, radios and real-device validation.

The backup list shows a green success indicator, and it looks as though you can finally replace the machine with confidence. But once the old host is shut down, you discover that the backup file is still on its hard drive, the decryption key was never saved elsewhere, and the external database was not backed up either.

The key distinction is this: a completed backup task proves only that the task completed the part it was responsible for. Whether the entire smart home can be restored also depends on the files, keys, dependent services, and hardware.

This article is based on Home Assistant’s official backup documentation, checked on November 10, 2026. It focuses mainly on Home Assistant OS and explicitly notes where Container differs. The recovery checklist is a recommendation; no migration or failure drill was performed on a user’s home devices.

First Check What the Backup Contains, Then Check Its Size

A Home Assistant OS backup can include settings, history, media, shared directories, and selected apps and app data, depending on what was selected when the backup was created.[1]

Successfully creating a file does not mean every item was selected. A small file might simply exclude media and history, or it might be missing data you actually need. A large file does not prove completeness either.

The official documentation also distinguishes between database locations: history in the default database can be included if selected; if the database runs as a Home Assistant OS app, that app needs to be included; if the database runs on another system, it is not in this backup and must be handled separately.[1]

The scope of a Home Assistant Container backup is not the same as that of an OS backup: the documentation primarily lists Home Assistant settings and optional history data. You cannot infer from that that other containers, volumes, and deployment files on the same host have also been packaged.

Before migrating, you can make an inventory of your actual dependencies:

Item What to check
Home Assistant configuration Whether the settings and automations you currently need are included
History and energy data Which database is in use and whether it is covered by the relevant backup
Services such as MQTT and databases Whether they run as OS apps, in another container, or on another machine
Media and shared directories Whether they were excluded and whether external files were saved separately
USB wireless adapters Whether the original adapter will be kept and how its device mapping will be restored
Network and external accounts Whether the new host can reach them and whether reauthentication is needed

This table is an inventory tool. Whether each item has actually been backed up should be determined from the installation method, configuration, and contents of the backup.

Neither the File Nor the Decryption Key Should Live Only on the Old Host

Home Assistant’s documentation says backups are encrypted by default. When setting up backups, you are prompted to download an emergency kit containing the key needed for recovery.[1]

If the backup you keep is encrypted, you need its corresponding key to restore it. If the key is stored only in the old machine’s downloads directory, it may be lost along with a failed drive. Saving the key without being able to retrieve the backup file will not allow a restore either.

Keep a readable copy outside the old host and safeguard the corresponding key. If you have changed keys, confirm which one an older backup uses; do not assume the emergency kit you can download now will decrypt every historical file.

Check whether a downloaded or transferred file is encrypted, too; do not judge by its extension alone. If a copy is a decrypted archive, treat it as a sensitive configuration file rather than casually uploading it to a public cloud drive.

The official recommendation is to keep a copy on another system and, ideally, another copy off-site.[1] For a household, getting to the point where “the old host cannot boot at all, but I can still find the file and key” is more meaningful than generating another file every day on the same drive.

Having Room for the Compressed File Does Not Mean There Is Room to Restore

A backup file is a compressed archive. Its size cannot be used directly to estimate how much space the new device needs.

Home Assistant’s migration documentation calls for checking the old system’s actual disk usage and allowing extra space on the target device. The recovery requirements also include the target system’s health and network connectivity.[1]

So a backup of a few hundred MB does not mean that any device with a few hundred MB free will suffice. The installation, restored database, apps, and media may take up much more space.

Also, “local control can work without the internet” does not mean “rebuilding the whole system requires no internet.” The official restore process reinstalls Home Assistant Core and apps, and the relevant Home Assistant OS recovery requirements include an internet connection.[1]

When preparing to migrate, check the target hardware, installation method, storage, and network needed for recovery together. If the interface is temporarily unavailable during a restore, do not immediately cut the power; check progress as the official instructions describe so you do not interrupt a normal rebuild.

The Dashboard May Be Back While the Devices Are Still Disconnected

After a restore, the original dashboards and automation configurations may appear, but the physical devices they depend on still need to work.

For example, the USB wireless adapter used by the old system may not be plugged into the new machine, or the virtual machine may not have the adapter mapped to it again. Device names remaining in the interface cannot replace a missing wireless interface.

Home Assistant’s official migration process reminds users to connect their existing controllers or wireless adapters during migration. If switching to a new built-in Zigbee radio, follow the Zigbee migration process rather than treating a configuration restore as a completed wireless-network migration.[1]

Network storage should also be checked for actual connectivity. Restoring the configuration for a shared path will not make a powered-off NAS accessible. If a battery-powered device is temporarily unavailable, wake it according to the device’s instructions and watch for reports rather than immediately deleting it from the system.

For services you deploy separately, verify that their data and connection settings have also been restored. This article does not claim that any Matter, Thread, or Zigbee hardware can be replaced seamlessly using only a Home Assistant backup.

Break “Restored Successfully” Into Observable Results

The backup system reports states such as creation success or failure and the time of the most recent successful backup. These are useful for detecting problems in the backup process, but they are not reports from a recovery drill.[2]

Consider setting an acceptance checklist for your own system:

  1. When the old host is unavailable, you can still retrieve the backup from another location.
  2. You can obtain the decryption key corresponding to that backup, along with the login credentials used when it was made.
  3. You can complete a restore in a suitable target environment and log in.
  4. You can confirm that critical automations, required history data, and apps are present.
  5. You can check one actual status report, one ordinary device-control action, and one low-risk automation.
  6. You can verify notifications and external dependencies, and record how long recovery actually took.

Schedule a drill during an acceptable maintenance window, using a spare environment or a clearly defined switchover plan. Do not allow a restored copy and the production instance to control the same devices or send duplicate notifications at the same time. If a drill checks configuration only in an isolated environment, record honestly that “control of real devices has not yet been verified.”

The restore process overwrites the selected parts. Before testing, retain an independent backup of the current system and establish conditions for rolling back. Do not overwrite the only working system just to prove that an old backup can be used.

A Backup Proves Its Value When You Need to Restore

Day to day, you can watch whether backups complete on schedule and whether remote copies are accessible. Before major changes or a migration, go further and confirm the scope of recovery, keys, and external dependencies.

The most useful thing to write down is not “I enabled automatic backups,” but: Which backup was most recently confirmed restorable, what does it contain, and which parts remain unverified?

That lets you base the next machine replacement on a record rather than a green icon.

Sources

  1. Home Assistant: Backups. Consulted for backup scope, external databases, encryption and emergency kits, separate storage, migration requirements, and post-restore checks; the page version identifier read was 2026.11.1.
  2. Home Assistant: Backup integration. Consulted for the meaning of states such as backup completion and failure events and the time of the most recent successful backup.

AI-generated concept illustration, not a test photograph or network topology diagram.

Discussion

Share a question or your experience. Comments appear after moderation. Your email is never displayed publicly. Comment privacy

Loading comments…