Cloud Backup: The 3-2-1 Rule

The 3-2-1 rule was established by a famous photographer, Peter Krogh, to remedy hard drive failures. It is considered an effective method of data backup, in both physical and virtual environments. When it comes to implementing a backup strategy, this is considered a great way. The principle is simple: 3 copies of data, 2 of which must be stored on different media, 1 of which must be kept outside the network.

3 copies of data

A copy of the data is never enough, especially since it is generally the original version. In this case, theft, loss or destruction of data can have noticeable consequences. Failure of an original or a copy is also possible, though likely over time.

In organizations with two backups, the probability of failure is evaluated at 1/10,000. But if the number of backups is increased to three, the probability of failure is evaluated at 1/1,000,000. One of these copies can take place in the cloud. The benefit of making at least three copies of the data is mathematically proven. Remember that the more copies of your data you have, the less likely you are to lose them.

2 copies stored on different media

A data storage medium (disk, tape, cloud) such as its environment (OS, applications, file formats, protocols, etc.) can fail. When backups are made on the same medium, if a failure occurs, all backups can be considered likely to have failed.

Whether it is Murphy’s law or normal breakdown, whether the mounts are purchased together and have the same mean time between failures (MTBF), it is quite common, after a support failure, to encounter a failure of the same support over- approximately the same. time. Hence the interest of storing backups on different media-storage servers, hard drives, encrypted USB keys, tape, cloud-and in different locations.

A backup copy kept offsite

A backup kept in the same place as others is not enough to protect the data. We saw this in the example of the fire at the OVH data center in Strasbourg: the backup servers were in the same room, or in a room located next to the production servers. As a result, the fire destroyed the original data and its copy (s).

If the backups are stored in the same place or connected to the same network, it is not sufficient to protect the data from the risks of theft, fire, water damage, damage or viral attack that would cause the data to be unusable. ito. The assigned backup media, for example in the cloud, should be kept outside the main site.

Towards 3-2-1-1-0

Latest development to date, the rule changes:

  • 3 copies;
  • 2 different supports or systems;
  • 1 offsite copy (usually online cloud);
  • 1 immutable offline copy (deduplication appliance, tape, cloud object storage);
  • 0 tracking through a proactive tracking layer that puts intelligence to monitor.

The cloud, 3-2-1 training, and DRaaS strategy

Physical separation between copies means storing them as much as possible. Building secondary data centers or private clouds for off-site data backups are costly options for construction and maintenance. Unless this is prevented by drastic or sovereign regulations, storing backups in the cloud is probably the most effective and cost-effective option today.

Beware, however, companies looking for a cloud backup solution also want to partner with a cloud service provider as a simplified way to secure off-site backups. But to also create a holistic approach to DRaaS, for the continued existence of virtualized environments.

If just protecting data is an important step, continuing to use it is the next step. Additionally, planning, creating, managing, testing, and maintaining backup and disaster recovery through 3-2-1-compliant locations is time and resource too.

The cloud can provide a “as a service” model that helps organizations save time, money, and resources in disaster backup and recovery. This model is based on a needs assessment measured by key metrics that meet RTO/RPO objectives. RTO (Recovery Time Objective) measures the time that critical IT resources may be unavailable before it significantly affects the business.

RPO (Recovery Point Objective), the loss of data that the organization can afford in the event of a disaster. RTO/RPO is a key element in defining the DRaaS strategy that comes with following the 3-2-1 rule. In general, we want them to be as small as possible.

Leave a Comment