Image-based backup creates a copy of an operating system and all the data associated with it, including the system state and application configurations. The backup is saved as a single file called an image. If hardened and used in production this is sometimes called a ‘golden image’.
The advantage to image-based backups is that all the information can be collected in a single pass, providing an updated bare-metal restore capability with each file-based backup.
Use Cases for image-based backups
Image-based backup products can perform backups online, perform dissimilar hardware restores, enable file-level restores, and recover servers remotely across wide area networks or local area networks.
They also enable backup images to be saved to a variety of different media. Many image-based backup products support encryption.
Because image-level backup applications use snapshots, all the data is backed up, including deleted files and empty disk blocks. To reduce the amount of data stored, some vendors incorporate data reduction technologies such as data deduplication.
VM snapshots and backups both retain VM data, but they satisfy different use cases. Admins can’t use snapshots as backups because snapshots rely on delta files to temporarily store VM data locally. Backups store VM data as a direct copy in a separate location, such as the cloud, which enables admins to restore the original VM for disaster recovery (DR) purposes.
The main distinction between backups and snapshots is that backups are independent, self-contained files that don’t require cross-file dependencies to restore a VM, whereas snapshots rely on dependent files for VM restoration.
VM snapshots enable admins to preserve VM data, such as files, configurations, memory and power settings at a specific point in time. For each VM, snapshots provide restore points that roll back a VM to when the system created the snapshot.
A system creates snapshots as incremental delta files — or change logs — that track any changes made since the previous snapshot. This process automatically occurs and provides a chain of connected delta files. Admins can then combine a specific snapshot with any archived snapshots and the original virtual disk to restore a VM.
Snapshot files are space-efficient because the system stores them as delta files and not exact VM replicas. Admins don’t need to wait for backup files to download from another location because the system can store snapshots alongside existing VM files.
But snapshots come with several challenges. They can quickly accumulate and take up valuable disk space, which can harm performance. The longer a system retains and accumulates snapshots, the more disk space admins lose. As a result, it isn’t practical to retain snapshots for any significant length of time, which makes them ineffective for long-term data protection.
Admins must also reconstruct snapshots to restore a VM. The system can’t restore a snapshot if a previous snapshot file or the original virtual disk is corrupted or unavailable. Admins can’t use snapshots to recover individual files; they must restore a snapshot in its entirety because the system deletes any files from previous snapshots.
Snapshots are useful for development and testing purposes. They make it easy to restore a VM to a point in time prior to a software upgrade, a patch release, new application deployment, configuration changes, driver installation or test runs. Snapshots also let admins restore a VM in the event of a cyberattack.
VM backups are exact copies of a VM that the system stores separately from primary VM files. Backup software, such as Veeam’s Backup & Replication software and Veritas’ NetBackup tool, copies a VM at specific intervals and moves those copies to another storage location. The backup software might also use data compression or deduplication to back up a VM.
Once the tool stores a VM backup, it remains isolated from the original VM and has no effect on primary VM operations. If the primary VM fails, the system retrieves the backup from the storage repository and restores the VM to the backup creation time and date.
Admins can store their VM backups on a variety of media types, including tape drives and the cloud, which lowers storage costs. Placing backups in a separate storage location makes it possible to isolate backups from the primary VM that can protect against natural disasters and cyberattacks. Backups aren’t dependent on other files, including the original virtual disk, which provides a more reliable VM protection option.
Before admins use VM backups, they must be aware that established backup methods often handle VMs inefficiently. If admins try to use a backup software that requires the system to install an agent on the guest OS, they should address potential performance or resource needs.
Because the virtualization layer sits between the guest OS and the physical hardware, the agent must go through the software layer to obtain VM data, which can result in unnecessary resource use. If a system simultaneously runs multiple backups, the added overhead can cause performance bottlenecks.
Admins also might need to pause VMs before the system can carry out backup operations to ensure they are in a consistent state and suitable to back up. Otherwise, some of the VM data might not be in a suitable state for restoration.
Backup processes can be resource-intensive and-time consuming, which is detrimental to host and VM performance, as well as application availability. Admins can perform fewer backups during a given period to retain resources, but a VM can quickly become out of date.
VM backups remain an essential part of any DR plan because backups can safeguard against data corruption, cyberattacks and natural disasters. VM backups are also beneficial when admins store backup files in locations separate from live VMs, which adds another layer of protection.
VM snapshots are valuable to any VM management strategy because they offer an efficient mechanism to quickly restore a VM to a specific point in time. Admins typically store snapshots locally and keep them for a relatively short duration; they rely on other files for VM restoration. Admins can store snapshots quickly, but they are unable to restore them if a dependent file is unavailable.
Backups are independent, self-contained files that admins can store indefinitely in a suitable location and recover without cross-file dependencies. Backups provide full copies of the VM, which makes them well suited for DR. The backup process incorporates snapshots to improve backup operations, but admins can’t only use snapshots as backups because they require dependent files.
