Flat backup is a method for using snapshots to protect data without using traditional backup software. Because this method of data protection doesn’t require traditional backup software, it is also sometimes referred to as a backup-less backup.
Advantages of Using Snapshots for Flat Backups
Flat backup leverages the snapshotting capabilities of the host storage system. Each time a storage system snapshots a dataset, it only copies the new or modified data, so after the initial snapshot, the process is likely to be faster than using backup software. Snapshots also require fewer resources, so they will have less of an effect on storage system performance.Content Continues Below
Traditional backup applications copy and save data in proprietary formats, so when the data is needed again, the backup application has to institute a restore process to access the backed up data and retrieve it in the data’s original format. If the backup dataset is voluminous, the process of restoring can be time-consuming.
Because snapshots are stored in the data’s native format, restoring a snapshot is fast.
Without the need to rely on specialized software, the primary advantages of a flat backup are its low cost and simplicity. Unlike a traditional backup, a flat backup does not require a backup server or a media server, which means it may be less expensive for storage administrators to implement. Flat backup simplifies the environment, reduces license fees, and improves recovery point objectives and recovery time objectives.
Limitations of Snapshot-based Backup
One of the drawbacks to relying on snapshots for data protection is that they are typically maintained on the same storage system that hosts the data being snapshotted. That’s okay for quick, limited restores where the original data was accidentally deleted, modified or was damaged. With flat backup, storage snapshots are copied to another system, preferably at another location. The storage snapshots still exist in their original location and native format, but the replicas serve as a backup if the originals get lost or damaged, or the system hosting those snapshots becomes unavailable.
Flat backups in the form of snapshots are proprietary, and not all storage vendors support them, although today most do. Because they are proprietary, snapshots are generally not “portable” and cannot be moved and accessed from one vendor’s storage gear to another’s. If a workload cannot be protected by a storage snapshot, it cannot be protected by a flat backup.
Bandwidth can become an issue if snapshots are replicated to a remote data center over a wide area network connection. To avoid a backlog, the flat backup mechanism must keep pace with snapshot creation. Bandwidth issues, however, may be mitigated after the initial snapshot replication as
Flat backup can also protect an application server if the storage hardware supports application-consistent snapshots. Application consistent means the snapshot system or the application that’s managing the process has insight into how specific applications handle data. For example, if the snapshotting process is not aware that a database system may not have fully committed a number of data modifications, there is the potential for data loss if the app and the snapshot aren’t in synch and mutually aware of the status of the app’s data. So if an organization is going to rely on snapshots for data protection, it’s essential that the snapshotting process is application consistent with all of the organization’s critical business applications.
One of the missing elements of snapshot-based data protection was an effective catalog that could track and index all snapshots as well as manage and schedule the process. Traditional backup applications have built-in catalogs that allow users to search for specific files and restores them without having to recall an entire dataset. Most backup application vendors have recognized the growing interest in flat backups and they have modified their applications so that in addition to managing their traditional backup processes, they can now also keep track of snapshots as well as manage the snapshotting and replication process.