Hurricane Dorian should make you take a hard look at your disaster recovery plans for critical operations. If the worst-case scenario were to occur, how resilient are your operations in the face of a catastrophic hurricane impacting nearly 50% percent of the U.S. Eastern Seaboard? With such a broad potential impact, I’m certain that many folks reading this article have 100% of their data and operations within Dorian’s cone! In fact, I’m writing this article while Dorian’s outer bands are making their way across Orlando.
Although implementing a comprehensive DR plan on the eve of the storm’s impact would be a fool’s errand, Dorian serves as an important reminder of how quickly a business can get wiped out if a disaster were to occur. Furthermore, with a major catastrophe fresh on everyone’s mind, the coming weeks present a critical opportunity to get buy-in from stakeholders to take the appropriate steps to ensure a major catastrophe like Dorian has little impact on operations. When designing your DR plan, here are some things to keep in mind:
Geographic Infrastructure Distribution
It’s as simple as “don’t put all your eggs in one basket.” Distributing your infrastructure continentally or even globally is a sure-fire way to protect your operations from anything less than the apocalypse. This is where public clouds such as AWS, Azure, and Google Cloud really shine. They provide cross-region replication for numerous forms of data storage, and mirroring capabilities with on-premise systems.
Cold vs. Warm DR Sites
How you implement DR can vary depending on use-case and budget. However, I recommend a fully operational recovery site, which mirrors the primary site. How warm (i.e. how much running infrastructure) you want to keep operational is up to you. Generally, I warn against cold sites because they often get neglected, and inevitably present problems when you try to awaken them from slumber. Warms sites ensure operators are used to working with the recovery site on a daily basis, and technical debt is resolved in the normal course of operations.
Workforce Impact Mitigation
Let’s not forget the people impacted by a major event such as Dorian. Prior to a hurricane, your workforce is going to have a major increase in personal responsibilities to prepare for the storm. In its aftermath, key personnel may be without power and internet for extended periods of time, and may even suffer serious property damage or personal losses. In the face of these events, it’s important to think about ways to pick up the slack with personnel outside the zone of impact, and/or provide capabilities to bring those impacted back online faster - for example, using cellular or satellite internet hotspots combined with battery backups to operate without the central power grid.
I’m a proud Florida native, and I’ve experienced nearly every hurricane to hit the state since the infamous Andrew in 1992. Living under a persistent threat of natural disasters means we go through the motions every year, personally and professionally, to prepare for the possibility of a hurricane. When it comes to your computing operations, take a page from Florida’s book and be ready to handle anything that comes your way.