Disaster recovery on and offsite

Adrian Moir explains his approach to disaster recovery for customers

Moir: Consider the number of customer locations

Deciding what data is important to your customer is vital when helping with their disaster recovery plans.

What are their business needs? Is email critical? What about archived data?

A good starting point is therefore classifying their existing data. It then becomes easier to define what needs protecting and how to do it. With different levels of protection available for different data categories, not all will become part of disaster recovery strategy although all data in a business must be protected.

More often than not, a one-size-fits-all offering will not be appropriate.

Acceptable recovery speed and level of data loss are two key criteria to be considered. Is it important to identify which services and data the customer will need first in the event of a disaster. Core business services can be quickly restored using offsite services.

Alternatively, the production systems could be placed offsite, while the disaster recovery location is back at the company office. This way, the manpower needed to implement the disaster recovery strategy is not in a remote datacentre with restricted access.

If an offsite location is required, determine what needs to be replicated to the customer’s disaster recovery site.

Looking at data movement within disaster recovery, one of the more popular methods is virtualisation, which allows the deployment of many standby servers in a smaller footprint and resultant cost reduction.

However, the real benefits of virtualisation start when it is deployed across multiple locations as the core systems available become easier to manage in terms of disaster recovery.

Most resellers planning disaster recovery for multiple sites will have to deal with the replication of the changing data sets. Many products enable data replication, but they do not always give the ability to catch issues around data corruption or logical data deletion.

I believe an out-of-band offering will give greater flexibility during recovery.

Finally, any sound disaster recovery strategy should also include regular checks to ensure the data is reaching the disaster recovery site, that it can be retrieved and that it can be validated.

Adrian Moir is technical director for EMEA at BakBone Software