[Fast Facts] 0-7 Disaster Recovery Levels

What is disaster recovery

Disaster recovery (DR) refers to the process by which an organization enables critical technical data, infrastructure, and systems to continue running after business disruptions caused by natural or man-made disasters. It is one aspect of business continuity.

It is essential to have a good disaster recovery plan in place, especially for IT departments. The most basic and common method of disaster recovery is taking regular backups, for example, virtual machine backup and recovery.

Based on the backup and recovery mechanism, disaster recovery is classified into 0-7 levels. I will explain them in this article.

0-7 Disaster recovery levels for enterprises

In the 1980s, the Share Technical Steering Committee and IBM built a 6-tier disaster recovery level model. As IT technology evolves, so has the model of disaster recovery levels.

Today in the IT industry, the commonly accepted disaster recovery levels have reached 7 tiers. The higher the disaster recovery level, the shorter recovery time required, but correspondingly, the higher of the recovery cost.

 0-7 tiers of disaster recovery explained:

Level 0: No off-site data

• Least expensive cost
• Recovery is only possible using on-site systems, no disaster recovery capability if your local server is destroyed.

This is how most organizations in the 1960s, 1970s, and 1980s used tape or disk drives to perform backups and restores.

Level 1: Physical backup with a cold site

• Transportation and storage space costs
• Recovery requires a long time

Physically remove tapes or other storage media that stores the data, and have them transferred and stored to an off-site facility by an external company.

Level 2: Physical backup with a hot site

• Transportation and standby site costs
• Reduced recovery time

Physically remove and transfer the storage media that stores the data, and restore them to a secondary site with the necessary hardware already installed.

Popular among large enterprises between the 1970s and 1980s, physical storage media was essential because of the slow network speed for data transfer at the time.

Level 3: Electronic vaulting

• Electronic transmit and standby site costs
• Recovery in hours

Data is transmitted electronically to an offsite site, without the need for physical transportation and storage space. Backup and recovery speeds depend on network bandwidth, which was increased during the 1980s and 1990s. To this day, this is still the most common data backup method.

Level 4: Point-in-time recovery

• More frequent data backup and recovery and standby site costs
• Recovery in minutes to hours

The two sites act as a backup for each other. More frequently recover image-level data backups to the offsite site with hardware and software already installed.

Level 5: Two-site commit/transaction integrity

• Continuous data transmission and standby site costs
• Recovery time in minutes

Run and maintain two sites simultaneously, and continuously transmit data in real time to another remote site to maintain data consistency. Substantial network bandwidth provided by the Internet is needed.

Level 6: Minimal to zero data loss

• Disk mirroring or data replication and standby site costs
• Instant recovery

With disk mirroring or data replication technologies, data files are backed up at the time of their creation and are continuously updated at a high frequency thereafter. Backed up data has the same age as primary data.

Commonly found in large enterprises with zero tolerance for data loss, such as banks or other telecommunication level enterprises. Today many companies choose cloud storage as another site.

Level 7: Recovery automation

• Most expensive
• Automatic instant recovery

The newest disaster recovery level.

Technologies such as artificial intelligence (AI) continuously monitor multiple aspects of data operations. When a possible abnormal situation was detected, it immediately analyzes the situation against a database of recovery rules, and triggers systems that already have critical data and applications for recovery in the shortest possible time.

This capability is typically available with cloud-based storage tools.

Practical use of disaster recovery plans for virtual machines

Virtualization is the core of the IT industry nowadays. Therefore, disaster recovery of VM data has become critical.

For small and medium-sized businesses, it is not cost-effective to create and maintain a standby site. The ability to automatically backup multiple virtual machines while running, cleanup old backups, and restore backups offsite is sufficient. And these features are provided by many professional backup tools. I recommend you try a free virtual machine backup software -- AOMEI Cyber Backup.

For large enterprises, backups may be not enough to meet their downtime or recovery time objective (RTO) requirements. To solve this issue, hypervisor vendors also provide VM failover solutions, and of course, with expensive licensing fees. The higher the disaster recovery level is, the higher it will cost.

For example, the VMware vSphere replication and Hyper-V live migrations functionalities on these most popular type 1 hypervisors.

Therefore, it is necessary to compare these features and develop a suitable disaster recovery plan.

Summary

There is no best disaster recovery plan, only the most suitable one. As you can see, the costs and recovery time varies for different disaster recovery levels.

To develop a suitable disaster recovery plan, firstly you need to be clear about your real needs. For example, how much downtime can your business afford, and how much data loss can your business tolerate?