Restoring is not always your best choice to recover data. This page introduces
alternatives and describes their strengths for various situations.

You can access your protected data in the following ways:

- [Clone](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/access-your-managed-data.md.txt#clone)
- [LiveClone](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/access-your-managed-data.md.txt#LiveClone)
- [Mount](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/access-your-managed-data.md.txt#mount)
- [Restore](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/access-your-managed-data.md.txt#restore)
- [Workflows](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/access-your-managed-data.md.txt#workflows)

## Clones

Clones are commonly used with SQL Server databases and VMware VMs.

### When to clone a Microsoft SQL Server database

Use the clone function to create an independent copy of a Microsoft SQL Server
database backups. The most common uses are the following:

- Application development and testing
- Data audit for compliance
- Data warehousing
- E-discovery
- User acceptance testing.

This can also be used like a restore, but to any server.

### When to clone VMware VMs

For VMware VMs, a clone copies backups of the VMDKs directly to a targeted
datastore before powering on the VM, at which point the VM is independent of
the backup/recovery appliance. This is like a restore to a different VM.

## LiveClones

A LiveClone can be used with all agent-based backups like databases and
filesystems.

The LiveClone is similar to the clone function, however, unlike a clone,
a LiveClone can be updated on demand or according to a schedule. When an
updated copy of the data is available, a LiveClone allows an independent copy
of a data to be mounted. This allows teams such as development and test to
ensure they are working on the latest set of data without having to manually
manage the data.

LiveClones enable you to do the following:

- **Refresh.** Create multiple, re-mountable copies of an active backup that can be refreshed from the most recent backup of the same application. See [Refresh a LiveClone](https://docs.cloud.google.com/backup-disaster-recovery/docs/access-data/liveclone#refresh).
- **PrepMount.** Mount the LiveClone backup to a specific host for a
  pre-processing operation (such as sanitizing data), even if some copies of
  this same LiveClone may already be mounted. You can prep-mount a LiveClone
  backup to scrub confidential or sensitive data.

  > [!NOTE]
  > **Note:** While the LiveClone backup is prep-mounted, it is busy. It cannot be deleted, refreshed, or mounted. If you mount application backups before prep-mounting another backup of the same application, the mounted backups continue to be usable.

- **PrepUnmount.** A prep-unmount operation on a LiveClone lets you decide
  if your changes are to be preserved or not. By default changes are tracked
  and preserved.

LiveClones provide a crucial building block in managing copy data lifecycle as
they combine the performance and independence of a full clone with the
efficiency and speed of incremental refreshes of data.

## Mounts

The Backup and DR mount function provides near instant access to
backups without moving data. These are several options for mounting data.

### VMware VM and agent based backups

The standard mount presents and makes application data available to a target
server as a file system, not as an application. This is useful if an application
is corrupt, lost, or if an application server is being replaced. In such cases
you can mount a backup and copy the application files from the mounted backup to
their original location on the application server.

For standard mounts:

- For VMware VMs ensure that NFS connectivity from the ESXi hosts to the backup/recovery appliance is available.
- For Compute Engine instances, ensure that the iSCSI port of the host where the active backups will be mounted is accessible to the backup/recovery appliance. Note that NFS is an option when mounting to instances running supported Linux operating systems.
- Mounting takes a copy of the selected active backup and mounts it to the selected host or VM.
- A mounted backup can be used directly from the mounted host.
- The original backup is never modified, even if the mounted virtual copy is modified or mounted and deleted.
- The mount operation is applicable for all agent based applications and VMware VMs.

Application aware mounts allow you to mount backups of databases as
virtual applications. This lets you quickly bring a database online without
having to actually move the data and without having to manually configure a new
instance of the database.
Application aware mounts are particularly useful in test and development
environments where multiple copies of a database must be quickly brought online.

Data presented as an application aware mount can be backed up like any other
application. Once the application aware mounted application data is backed up,
it too can be can be mounted as an application aware mount.
The backup, application mount, backup sequence can be repeated to any depth.
By default, the sequence is restricted to five generations of the original
database.

Mount and migrate lets you restore an application with near-zero downtime
by first mounting it locally, and then migrating it to the original location or
to a new location. Users have normal access to the application while it is
mounted, and the migration step is very fast.

When performing a mount of a VMware VM or agent based backup from OnVault you
can control how much to optimize for performance versus storage consumption,
by selecting the following:

- **Storage-optimized.** This only keep writes in the local snapshot pool
  (writes are always kept locally).

- **Balanced.** These blocks are read (from object storage) or written (to
  local snapshot pool) and kept in the snapshot pool, to serve as a cache for
  future reads.

- **Performance-optimized.** This brings the entire backup to the local snapshot
  pool, in the background. Reads will become faster as more of the backup is
  available locally.

- **Max performance.** The entire backup is rehydrated into the snapshot pool
  first, prior to the mount. This means that the host always works against local
  storage only.

### Compute Engine instance

Compute Engine instance mounts create a new Persistent Disk from backup and
then either assign it to an existing Compute Engine instance or to a new
Compute Engine instance, which is created at the same time.
Unlike VMware VM or agent based backups, this does involve the movement of
data from where the backups are stored (in Google Cloud) to
Persistent Disk.

## Restores

The restore function reverts the production data to a specified point in time.
Typically restore operations are performed to restore an application to a valid
state after data corruption. The amount of time required to complete a restore
operation depends on the amount of data involved.

## Workflows

A workflow automates access to copy data. While backup plan policy templates
govern the automated capture of production data, workflows automate the access
to this data.

Steps are defined within a workflow to perform a series of tasks on a schedule
or on demand. This includes creating and refreshing LiveClones, data masking,
persistent mounts, and non-persistent processing mounts for tasks such as
database integrity checks, and ETL loads. Workflows are also used by
administrators to provide simplified and secured self-service data access to end
users such as database administrators and application developers.

## The Backup and DR getting started guide

This page is one in a series of pages selected to introduce Backup and DR.
The other topics are as follows:

- [Introduce the appliance management console](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/introduction)
- [Introduce Backup and DR data protection backup plans](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/backup-plan)
- [Introduce Backup and DR data capture](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/data-backup-overview.md.txt)
- [Different ways to access your protected data (this page)](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/access-your-managed-data.md.txt)
- [Introduce the Backup and DR agent](https://docs.cloud.google.com/backup-disaster-recovery/docs/concepts/about-connector)