Today, backups need to satisfy plenty of requirements to ensure the possibility to recover data in various situations. To make sure your backup infrastructure covers the most important backup requirements, I suggest using the 3-2-1-1 strategy. The 3-2-1-1 strategy states the following: - Three independent full data backups. - Two backup-media devices that contain at least one current full data backup. - One current full data backup stored at a separate location from the main infrastructure. - One current full data backup, that is either stored offline or is immutable. To help with this situation, you can see five Backup strategies that can be combined in various ways to accomplish the 3-2-1-1 requirements. All backup strategies follow a strict standard and have a specific function, do not abuse them for purposes that aren't mentioned below.
- Three different copies of data
- Two different media
- One offsite copy
- One offline/immutable copy
Cloudbased Revision Control Backup
Revision Control Backup is a backup strategy aiming to restore smaller data sets quickly and accurately.
Regular backups capture data changes, allowing precise restoration to specific past versions of the data.
What it is
Flexible short-term backup that retains the most recent backups for specific days.
Disaster recovery backup, since it's stored in a location different from the main infrastructure, can indeed serve as a disaster recovery solution.
What it is not
Revision control system for end users. It is not designed to address or mitigate user mistakes, but system or software failures!
Target
The cloudbased RC strategy uses S3 Buckets as backup target.
Backups are not set to immutable due to the need for flexibility.
This approach allows for the creation of manual backups in situations where backups need to be added or removed manually (e.g. before a major update).
Retention
The retention policy used in the cloudbased RC strategy is the GFS (Grandfather-Father-Son) policy.
As configuration, enable GFS and define a day to create synthetic full backups.
The policy should at least keep 14 daily incrementals, 4 weekly backups and 2 monthly backups.
Depending on your needs, daily, weekly or monthly restore-points can be increased, I do not recommend to create yearly backups with this strategy
Local Revision Control Backup
Revision Control Backup is a backup strategy aiming to restore smaller data sets quickly and accurately.
Regular backups capture data changes, allowing precise restoration to specific past versions of the data.
Since this backup targets a NAS device, it's typically more affordable than cloud solutions.
What it is
Flexible short-term backup that retains the most recent backups for specific days.
What it is not
Revision control system for end users. It is not designed to address or mitigate user mistakes, but system or software failures!
Disaster recovery backup, although this backup is located in the same location as the rest of the infrastructure, it shouldn't be considered a true disaster recovery solution.
Target
The local RC strategy uses a NAS device as backup target.
Usually the target is connected to the backup-instance via SMB-Share.
Retention
The retention policy used in the local RC strategy is the GFS (Grandfather-Father-Son) policy.
As configuration, enable GFS and define a day to create synthetic full backups.
The policy should at least keep 14 daily incrementals, 4 weekly backups and 2 monthly backups.
Depending on your needs, daily, weekly or monthly restore-points can be increased, I do not recommend to create yearly backups with this strategy
Cloudbased Long Term Archive Backup
The Long Term Archive backup strategy is designed to retain immutable backups over extended durations.
Both the LTA and RC strategies capture data changes. However, the LTA emphasizes long-term storage and creates immutable backups, making it less flexible than the RC backups.
What it is
Long-term archive backup, designed to store backups over an extended time periods.
What it is not
Flexible backup that can be used to manually add and remove restore points.
Archive storage, used by end users to retain documents, such as legal papers, that require preservation for specific time periods.
Target
The cloudbased LTA strategy uses S3 Buckets as backup target.
Backups are set to immutable to ensure their integrity over time.
Retention
The retention policy used in the cloudbased LTA strategy is the GFS (Grandfather-Father-Son) policy.
As configuration, enable GFS and define a day to create synthetic full backups.
The policy should at least keep 14 daily incrementals, 2 weekly backups and 5 yearly backups.
Depending on your needs, daily, weekly, monthly or (especially) yearly restore-points can be increased.
Cloudbased Disaster Recovery Backup
The Disaster Recovery backup strategy aims to swiftly restore a recent and consistent backup.
This DR approach stores backups for a short period of time, with the main purpose being to ensure infrastructure recovery during emergencies.
What it is
Consistent backup that provides the fastest possible recovery in the event of a disaster.
Immutable backup that cannot be encrypted during a ransomware attack.
What it is not
Anything else!
Target
The cloudbased DR strategy uses S3 Buckets as backup target.
Backups are made immutable to safeguard against encryption during a ransomware attack.
Retention
The retention policy used in the cloudbased DR strategy is the GFS (Grandfather-Father-Son) policy.
As configuration, enable GFS and define two days to create synthetic full backups (e.g. tuesday and friday).
The policy should at least keep 14 daily incrementals and 2 weekly backups.
Depending on your needs, daily or weekly restore-points can be increased, I do not recommend to create monthly or yearly backups with this strategy
Local Disaster Recovery Backup
The Disaster Recovery backup strategy aims to swiftly restore a recent and consistent backup.
This DR approach stores backups for a short period of time, with the main purpose being to ensure infrastructure recovery during emergencies.
In the local DR strategy, the concept involves rotating between at least two NAS devices weekly. The unused NAS device is typically stored offsite and kept offline.
What it is
Consistent backup that provides the fastest possible recovery in the event of a disaster.
Offline backup that cannot be encrypted during a ransomware attack.
What it is not
Anything else!
Target
The local RC strategy uses at least two NAS devices as backup target.
NAS devices need to be rotated once a week. The inactive device must be kept offline in an offsite location.
Usually the target is connected to the backup-instance via SMB-Share.
Retention
The retention policy used in the local DR strategy is the forward incremental policy.
As configuration, define two days to create synthetic full backups (e.g. tuesday and friday).
The policy should at least keep 7 restore points.
For this retention make sure you use restore points instead of days.