Disaster recovery capabilities are now an accepted and widespread requirement for businesses to have in place but there are three important areas that are often overlooked which can reduce their effectiveness, says Peter Godden.
For many organizations, disaster recovery (DR) is becoming part of their everyday reality with Gartner’s 2015 Business Continuity survey reporting that 72 percent of organizations had used their DR plan. That such a huge percentage of organizations used their DR plan should serve as a wake up call for anyone without a DR plan in place. Far too often, DR planning happens in the wake of a disaster when it is too late to ensure business continuity. The best plans should be put in place long before data is lost, business is halted and the organization suffers.
Even with a plan in place, business continuity might not be guaranteed. What happens if the DR site goes down or the DR plan isn’t executed well because everyone has gone into crisis mode? Whether you are creating a DR plan for the first time or updating an existing one, there are three key considerations to include to ensure the plan is robust, effective and allows true recovery to take place:
A clearly documented failover process is a must for a DR plan to be successful, but an automated one is better. Manual failover processes have many steps and each manual step is an opportunity for a mistake to be made. If you are executing a production failover, you are in crisis and that can cause people to panic, elevating the probability of human error. An automated failover ensures consistency and repeatability. Automation eliminates the risk of mistakes while ensuring fast, seamless and predictable recovery.
Don’t just failover, failback
Once organizations have an automated and well-documented failover process, they should create an automated and well-documented failback process. During downtime companies want business continuity, but once they are back up and running, they want to be able to transfer the data from the DR site back to the production site.
Many organizations will not execute a failover because they do not have a documented failback process and/or they know they cannot successfully failback. This leaves them more vulnerable to a variety of data centre / center disruptions that can significantly hurt business operations. The failback should not be more cumbersome than the original failover. To be completely confident it will work, organizations should fully test production failover and failback.
Have a DR plan for your DR
What do you do if the DR site is down? There are some instances where DR sites – even if located far apart – can experience a disaster recovery failure. In 2003, for example, there was a huge blackout spanning from the northeast to other parts of the US and eastern Canada. Someone in Massachusetts who had located their DR site “far away” in Ohio still couldn’t recover due to the wide range of the power outage. This is why it is critical to have Plan A, Plan B and Plan C. For example, recovering to a public cloud provider can offer greater geographical diversity to avoid situations like this.
Downtime causes damage to brand and company reputation, lets down employees and customers – causing frustration and productivity issues – and can seriously affect revenue. An automated plan that takes over in a crisis to failover and failback smoothly can prevent downtime and disaster. With an additional plan to protect from disasters at the DR site, you can sleep easy knowing everything has been taken care of.
Peter Godden is VP, EMEA at Zerto.