A few weeks ago, Hurricane Matthew ripped a path of destruction through Haiti, Cuba, the Bahamas, and the southeastern coast of the USA.
The USA was spared the worst of the storm—Matthew didn’t make landfall until reaching the South Carolina coast as a Cat 1 hurricane.
However, on its way to the SC coastline, the eye of the storm passed parallel to the Florida and Georgia coastlines.
As it passed by, hurricane-force winds emanating from the eyewall of the storm lashed the coastal and central portions of Florida. Reports from The Weather Channel’s website highlight that some coastal areas experienced wind gusts in excess of 80-90 mph.
The heavy winds and rain caused severe damage to the state’s electrical infrastructure, causing power outages for hundreds of thousands of Florida’s residents—many of whom had to wait for days to get their power restored.
ZDNet did an article after the hurricane was over highlighting some of the similarities between electric companies and cloud service providers as distributed service companies upon whom many customers rely—and the effects of disruptions in that service caused by natural disasters such as Matthew.
This article begs the question: What can we learn about disaster recovery (DR) from the events of Hurricane Matthew?
Lesson 1: Don’t Put All of Your Eggs in One Basket
Distributed services are extremely convenient—until something happens to the central location from which services originate, that is.
While operating from one central location is easier to build, manage, and maintain, than having multiple locations, disasters such as hurricanes make it all too easy for services to be disrupted. This “all eggs in one basket” distribution model creates too many single points of failure between the provider and the customer.
If one transformer blows between the power plant and the customer, then power is lost until a technician can be dispatched to fix or replace it.
The need for redundancy in case of failure holds true for cloud service providers as well. When delivering services via the cloud, providers have to maintain geographically-diverse redundant systems in place so that single points of failure are eliminated.
Disaster recovery plans should use multiple facilities, and each facility should have multiple connections so that no one blown circuit or compromised storage drive brings down the service.
Lesson 2: Disaster Response Requires Preparation and Swift Action
Hurricane Matthew was on weather maps for weeks before brushing up against the Americas. States of emergency were declared well ahead of the hurricane passing along the Florida coast, and evacuation alerts were generated days before it struck.
As noted in the ZDNet article, the author had spoken to “some of the guys working the lines, and they told me they’d been flown in from out of state before the storm.” In other words, the electric company had recognized that the hurricane was going to be severe enough to cause significant damage to the state’s electrical infrastructure, and prepared a special emergency response team accordingly.
While cloud service providers don’t usually get weeks of warning before a “disaster” strikes a data center, their disaster recovery plans benefit from some forethought and preparation.
Lesson 3: Maintenance is Key for Mitigating Disasters
Part of mitigating a disaster event and being prepared is maintaining key aspects of the infrastructure in top condition. Infrastructure that isn’t properly maintained is simply more prone to failure in a crisis situation.
This was highlighted in ZDNet’s article, when the author described vine-covered transformer behind his house: “notice the overgrowth? If a branch crosses over the connectors, that transformer will either spark or explode. It’s already exploded once.”
Failure to maintain infrastructure for a disaster recovery plan can have disastrous results as well. A redundant system that isn’t maintained and fails when it’s most needed really isn’t providing the redundancy a true DR solution needs.
Be Prepared for When Disaster Strikes
When Hurricane Matthew brushed the Florida coast, hundreds of thousands of U.S. citizens were left without power and other basic services for the better part of half a week. Even with specific preparation beforehand and weeks of warning, Matthew still caused millions of dollars in damage throughout Florida, Georgia, and the Carolinas.
The disasters that can bring down a business’ infrastructure or cause data loss often happen completely without warning. From hackers attacking your production environment, to accidental deletions of mission-critical files, to unexpected power outages, there are any number of manmade disasters that can befall a business.
Having a disaster recovery plan in place that is ready to go on a moment’s notice is a must in a world where natural and manmade disasters can cause disruptions in service. Find out how you can get powerful DR solutions from a best-in-class provider with WHOA.com today!