# A recovery rehearsal you can repeat.

> Use a practical restore checklist: choose a snapshot, retrieve its key, isolate the target, restore the data, validate the application and record recovery time.

Canonical page: https://www.calmbackup.com/guides/restore-checklist

A practical checklist for choosing a snapshot, preparing a safe destination, validating the application and recording the result.

[Start free trial](https://app.calmbackup.com/register)

## Before you touch the restore command.

Choose what this rehearsal needs to prove: recovery after losing a server, recovery from an earlier snapshot, or simply that a new backup setup works. Write down the expected outcome so a successful command is not the only pass criterion.

1.  ### Identify the snapshot
    
    Record the source, backup identifier and timestamp. Choose data that lets you check meaningful application behavior.
    
2.  ### Recover the key
    
    Confirm you can access the encryption key without relying on the source machine. Use the key associated with this archive.
    
3.  ### Prepare an isolated destination
    
    Install the database tools and application dependencies. Confirm the target database and file paths; separate the environment from production traffic and outbound integrations.
    
4.  ### Review the configuration
    
    Use an explicit recovery configuration. Have another operator check the destination when the rehearsal involves valuable existing data.
    

## Run the restore. Keep the evidence.

The CLI obtains the archive, authenticates and decrypts it, extracts the contents and invokes the database restore. Record the output and any warnings. Stop to understand unexpected errors rather than treating a partially completed import as success.

```
calmbackup --config /path/to/recovery.yaml restore <backup-id>
```

**The target can be overwritten**

Review database and directory destinations before running the command. An isolated rehearsal should never use the production configuration by accident.

[Open the command reference](https://www.calmbackup.com/docs/commands)

## Check what your users would notice.

Assign an owner to any problem found during the rehearsal. Repeat the relevant steps after fixing it, and schedule the next exercise around changes to the application or infrastructure.

1.  ### Inspect the important data
    
    Check representative records and their relationships. Include older records as well as recent ones from before the snapshot.
    
2.  ### Open the associated files
    
    Check uploads and generated documents. Verify ownership and permissions as well as the presence of filenames.
    
3.  ### Exercise the application
    
    Run representative reads and writes in the isolated environment. Inspect queues or scheduled operations without triggering production side effects.
    
4.  ### Record the full recovery time
    
    Include obtaining credentials, preparing the environment, downloading, importing and validating. Keep the result and any missing steps in the runbook.
    

[Understand recovery objectives](https://www.calmbackup.com/guides/rpo-rto)
