[Set up a backup vault](https://docs.cloud.google.com/backup-disaster-recovery/docs/cloud-console/backup-vault-create.md.txt)

The backup vault is a Google-managed regional resource that provides
isolated, immutable, and indelible storage for backups. It organizes data into a
hierarchy of *vaults* , *data sources* , and *backups*, protecting data against
accidental or malicious deletion through enforced retention periods and
project-level identity isolation.

This document explains how backup vaults work, including the
hierarchical resource model, supported workloads, backup models, location
compatibility, availability, and naming requirements.

## Resource hierarchy

Data within the Backup and DR is organized into a three-tiered hierarchy
within the backup vault:

- **Backup vault**: a top-level container that enforces global
  minimum retention policies.

- **Data source**: a child resource representing a specific protected entity
  (for example a Compute Engine instance). The system creates this
  automatically upon the first backup.

- **Backup**: a child resource of a data source representing a discrete backup
  with a specific point-in-time recovery point.

The following diagram shows the backup vault resource model:
![The backup vault resource model.](https://docs.cloud.google.com/static/backup-disaster-recovery/docs/images/concepts/resource-model.png) Backup vault resource model.

## Limitations

- This page describes backup vaults for resources managed through
  the **Google Cloud console** (such as Compute Engine or Cloud SQL). If
  you are using the **appliance management console** for workloads like Oracle
  or Google Cloud VMware Engine, see [Overview of backup vaults managed from the appliance management console](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/backup-vault-with-ba).

- AlloyDB clusters and Filestore
  instances in backup vaults aren't supported for multi-regions.

- Backup and DR doesn't impose any restriction on the compatible
  destination locations when restoring a workload from a
  backup vault backup.

- You can incur network transfer fees depending on the locations of your
  backup vault and source workload. For more information, see
  [Backup and DR pricing](https://cloud.google.com/backup-disaster-recovery/pricing).

## Supported resources

| Workload type | Management |
|---|---|
| Compute Engine instance | Google Cloud console |
| Compute Engine disk | Google Cloud console |
| Filestore instance | Google Cloud console |
| Cloud SQL instance | Google Cloud console |
| AlloyDB cluster | Google Cloud console |
| Google Cloud VMware Engine, Oracle database, and SQL Server database | Appliance management console |

## Backup models for resources

|   | Centralized model | Decentralized model |
|---|---|---|
| **Google Cloud console** | Consolidates backup management by creating backup vaults and backup plans in a central administrator project. Administrators can use these centrally managed plans to protect resources across multiple service projects, or use IAM permissions to delegate backup plan access to application owners. | Isolates backup management by creating a separate backup vault and backup plan within each project. This approach is ideal for decentralized organizations where individual application teams are responsible for backing up their own resources. |
| **Appliance management console** | Consolidates backup management by deploying the appliance management console and creating backup vaults in a central administrator project. Administrators configure backup policies within the central management console to protect resources, such as Google Cloud VMware Engine VMs, across multiple service projects. | Isolates backup management by deploying a separate appliance management console and backup vault for each project or line of business. This approach is ideal for decentralized organizations where backup management responsibilities are split across multiple teams. |

## Backup vault supported locations

You can create a backup vault in the same region as the source
workload (regional), in a different region from the source workload
(cross-regional), or across multiple regions (multi-regional).

### Supported regions and cross-regions

You can create backup vaults in the following regions and
cross-regions:

| Geographic Area | Region Name | Region Description |   |
|---|---|---|---|
| **North America** |   |   |   |
|   | `northamerica-northeast1` ^\*^ | Montréal | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `northamerica-northeast2` | Toronto | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `us-central1` | Iowa | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `us-east1` | South Carolina |   |
|   | `us-east4` | Northern Virginia |   |
|   | `us-east5` | Columbus |   |
|   | `us-south1` | Dallas | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `us-west1` | Oregon | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `us-west2` | Los Angeles |   |
|   | `us-west3` | Salt Lake City |   |
|   | `us-west4` | Las Vegas |   |
|   | `northamerica-south1` ^\*^ | Querétaro |   |
| **South America** |   |   |   |
|   | `southamerica-east1` | São Paulo | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `southamerica-west1` | Santiago | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
| **Europe** |   |   |   |
|   | `europe-central2` | Warsaw | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-north1` | Finland | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-north2` | Stockholm | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-southwest1` | Madrid | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-west1` | Belgium | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-west2` | London | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-west3` | Frankfurt |   |
|   | `europe-west4` | Netherlands | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-west6` | Zürich | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-west8` | Milan | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-west9` | Paris | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
|   | `europe-west10` | Berlin |   |
|   | `europe-west12` | Turin | ![leaf icon](https://cloud.google.com/sustainability/region-carbon/gleaf.svg) [Low CO~2~](https://cloud.google.com/sustainability/region-carbon#region-picker) |
| **Middle East** |   |   |   |
|   | `me-central1` | Doha |   |
|   | `me-central2` | Dammam |   |
|   | `me-west1` | Israel |   |
| **Africa** |   |   |   |
|   | `africa-south1` | Johannesburg |   |
| **Asia Pacific** |   |   |   |
|   | `asia-east1` | Taiwan |   |
|   | `asia-east2` | Hong Kong |   |
|   | `asia-northeast1` | Tokyo |   |
|   | `asia-northeast2` ^\*^ | Osaka |   |
|   | `asia-northeast3` | Seoul |   |
|   | `asia-southeast1` | Singapore |   |
|   | `asia-southeast2` | Jakarta |   |
|   | `australia-southeast1` | Sydney |   |
|   | `australia-southeast2` | Melbourne |   |
| **India** |   |   |   |
|   | `asia-south1` | Mumbai |   |
|   | `asia-south2` | Delhi |   |

^\*^ Querétaro (northamerica-south1), Montréal (northamerica-northeast1),
and Osaka (asia-northeast2) don't support zone separation. This means the multiple
zones within each of these regions may not be located in physically separate
data center campuses. Consequently, a single, localized physical disaster event
could potentially impact multiple zones within the same region, increasing the
risk of data loss compared to regions that support zone separation.

### Supported multi-regions

You can create backup vaults in the following multi-regions:

| Multi-region name | Description |
|---|---|
| `ASIA` | Data centers in Asia |
| `EU` | Data centers in the European Union |
| `US` | Data centers in the United States |

### Workload location compatibility

The following table describes the compatible backup vault locations
for each supported workload, when using regional and cross-regional
backup vaults. Note that backup plans in the Google Cloud console must be
created in the same region as the source workload.

| Workload | The backup vault must be in the same region as the source workload | Regional support | Multi-regional support | Cross-regional support |
|---|---|---|---|---|
| Compute Engine instance | No |   |   |   |
| Compute Engine disk | No |   |   |   |
| Cloud SQL instance | No |   |   |   |
| AlloyDB cluster | Yes |   |   |   |
| Filestore instance | No |   |   |   |
| Google Cloud VMware Engine, Oracle database, and SQL Server database | No |   |   |   |

### Multi-region compatibility

The following requirements must be met to use multi-regions:

- If a workload supports multi-region backup vaults, the source
  workload location must be compatible with the multi-region backup vault
  location.

- You can only backup up resources in regions that share the same prefix. For
  example, resources in regions with the `asia` prefix are only allowed to be
  backed up to the `asia` multi-region.

The following table describes the compatible backup vault locations
for each supported workload when using multi-region backup vaults:

| Workload type | Supports use of multi-region backup vaults? | Supported backup vault multi-regions |
|---|---|---|
| Compute Engine instance |   | `asia`, `eu`, `us` |
| Compute Engine disk |   | `asia`, `eu`, `us` |
| Filestore instance |   | N/A |
| Cloud SQL instance |   | `asia`, `eu`, `us` |
| AlloyDB cluster |   | N/A |
| Google Cloud VMware Engine, Oracle database, and SQL Server database |   | N/A |

### Availability

Backup vaults created in regional and cross-regional locations
provide resilience against a single-zone outage. Backup data is stored
redundantly in at least two separate zones.

Backup vaults created in multi-region locations provide
resilience against a single-region outage. Backup data is stored redundantly
in at least two separate regions.

### Compare multi-region and cross-region backup vaults

| Criteria | Multi-region backups | Cross-region backups |
|---|---|---|
| Backup creation | Automated by Google across two regions within a continent. | Autonomy to explicitly define a region to create backup. |
| Use case | High availability \& operational simplicity | Strict compliance, data residency laws, or targeted disaster recovery sites, within and outside source regional boundaries. |
| Management | Low overhead. Single vault, automated balancing | Medium overhead. Requires setting up specific target pairing. |
| Customer-managed encryption keys (CMEK) | Multi-region backup vault must use CMEK from the same region as the backup vault. | Cross-region backup vault must use CMEK from the same region as backup vault. |
| Cost implications | Multi-regional upload and download charge, [where applicable](https://cloud.google.com/backup-disaster-recovery/pricing?e=0#multi-regional-upload-download-charges). Backup storage charge for multi-region vault. Management charge. | Inter-region data transfer charges. Backup storage charge. Management charge. |
| Supported workloads | - Compute Engine instances - Compute Engine disks - Cloud SQL instances | - Compute Engine instances - Compute Engine disks - Filestore instances - Cloud SQL instances |

## Backup vault names

Your backup vault names must meet the following requirements:

- Backup vault names can only contain lowercase letters, numeric
  characters, and hyphens (`-`). Spaces aren't allowed.

- Backup vault names must start and end with a number or letter.

- Backup vault names must contain 3-63 characters. Names containing
  periods can contain up to 222 characters, but each period-separated component
  can be no longer than 63 characters.

- Backup vault names cannot be represented as an IP address in
  dotted-decimal notation. For example, `192.0.2.255`.

## Prevent backup deletion

To protect your data from accidental or malicious deletion, you (the administrator) configure an enforced retention period for your backup vault. Once configured, backups are completely immutable---they cannot be manually deleted by you, any other user, or Google until your specified duration has passed.

While enforced retention prevents the deletion of individual backups within a vault, project liens add an additional layer of protection by blocking the deletion of the project itself. For more information, see [Protect projects with liens](https://docs.cloud.google.com/backup-disaster-recovery/docs/configuration/project-liens.md.txt).

> [!NOTE]
> **Note:** Because backups cannot be deleted during enforced retention, they will incur standard storage charges for the full duration. Before configuring enforced retention, review the cost and compliance implications described in [Resource hierarchy](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/backup-vault.md.txt#resource-hierarchy).

You control enforced retention using three main vault settings:

- Minimum enforced retention: When creating a vault, you must set a baseline retention period (between 0 days and 99 years). This setting creates a mandatory vault-wide floor:

No backup can be deleted before this time passes.
Any backup plan storing data in this vault must have a backup retention period equal to or longer than this minimum.

- Inherit retention from backup rule: Instead of relying solely on the vault's baseline minimum, you can configure the vault to adopt the exact retention period defined in your specific backup plans. You must enable this setting during vault creation.

  If your vault's minimum is 3 days, but your backup plan's retention is 7 days, the backup is locked for the full 7 days.
  Result: Manual deletion is completely prevented. Backups are only deleted automatically once the backup plan's specific duration expires.
- Lock the enforced retention period: For strict compliance requirements, you can permanently lock your minimum enforced retention configuration.

When setting a lock, you select an effective date.
Before the effective date, you can still adjust the minimum retention period up or down to correct mistakes.
After the effective date, no one (not even a Project Owner) can decrease the retention period. You are only permitted to increase it.

> [!IMPORTANT]
> **Important:** Beginning November 1, 2026, Backup and DR automatically applies a project-level lien to any project containing a backup vault with backups protected by enforced retention.

## Backup vault access restriction

A backup vault's access restrictions setting lets you control the
sources from which data can be backed up into or restored from a
backup vault. This setting determines the types of resources that you
can store in a backup vault.

You can select one of the following access restriction settings for a
backup vault. Note that this setting is permanent and can't be changed.

- **Restrict access to current organization**: backup and restore operations
  are only supported within your current organization. This selection makes
  the backup vault compatible with resources managed through
  Google Cloud console, such as Compute Engine instances, but not with resources
  managed through the appliance management console.

- **Restrict access to current project**: backup and restore operations are
  only supported within your current project. This selection makes the
  backup vault compatible with resources managed through
  Google Cloud console (for example, Compute Engine instances), but not with resources
  managed through the appliance management console.

- **Restrict access to current organization \& unrestricted access for backup appliances**:
  for resources managed through Google Cloud console, backup and restore operations
  are only supported within your current organization. Resources managed through
  the appliance management console (for example, Google Cloud VMware Engine VMs) are also
  supported, but backup and restore operations for those resources are not
  restricted to your current organization. This selection makes the
  backup vault compatible with resources managed through
  Google Cloud console and with resources managed through the appliance management console.

- **Allow unrestricted access** : allows backup and restore operations to or
  from *any* project or organization. This selection makes the backup vault
  compatible with resources managed through Google Cloud console and with
  resources managed through the appliance management console.

## Encryption

By default, Google Cloud automatically encrypts data when it's at rest
using [Google-owned and Google-managed encryption keys](https://docs.cloud.google.com/docs/security/encryption/default-encryption.md.txt).
If you have specific compliance or regulatory requirements related to the keys
that protect your data, you can use customer-managed encryption keys (CMEK)
for your backups. See [Customer-managed encryption keys (CMEK)](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/cmek.md.txt).

## What's next

- [Automate protection of Compute Engine instances](https://docs.cloud.google.com/backup-disaster-recovery/docs/create-plan/dynamic-protection.md.txt)
- [Create and manage a backup vault in the Google Cloud console](https://docs.cloud.google.com/backup-disaster-recovery/docs/cloud-console/backup-vault-create.md.txt)
- [Back up disks to a backup vault](https://docs.cloud.google.com/backup-disaster-recovery/docs/cloud-console/compute/disk-backup.md.txt)