What We Learned from Facebook's Historic Downtime

      Posted by Matthew Fall on Mar 26, 2019 6:05:00 AM

      You probably heard Facebook was down earlier this month. Or, you’ve completely cut yourself out from the world of social media (props to you). Starting around noon last Wednesday, March 13th, almost 2.3 billion people were unable to access Facebook, its apps, and services due to worldwide outages. The downtime lasted about 14 hours, impacting Facebook, Instagram, WhatsApp, Messenger, and other Facebook-owned services. This unplanned downtime had serious ramifications, like a 1.8% drop in the stock price of Facebook, negative publicity, and negative impacts on businesses that use Facebook’s services to be productive. This includes advertisers and people who pay for Facebook services.Candid image of a group with succesful business people caught in an animated brainstorming meeting

      What Caused the Downtime? 

      When the platform first went down, it was rumored that the outage was caused by a DDoS (denial-of-service) attack. DDoS is a cyber-attack that floods a company’s network with traffic- so much traffic that it jams the network and makes it unavailable for intended use. DDoS is just one type of cyber-attack among many others, like ransomware, viruses, phishing attempts, and malware. 

      DDoS ultimately wasn’t the issue that caused Facebook to go down. Instead, the issue was with a server reconfiguration change that triggered a series of additional problems for Facebook. As a result, some people were unable to post comments or access the site altogether.  

      Why this could happen to your business 

      What seemed like a routine server reconfiguration turned into 14+ hours of downtime and lost revenue for Facebook. The more shocking part is that Facebook employs some of the best IT professionals in the world- the type of professionals that most companies can’t afford to employ. More importantly, it shows that outages do occur and without a plan in place to combat unplanned downtime, your business could incur serious loss.

      Unplanned downtime could be a result of something as simple as a server reconfiguration or human error, or even due to a more complex issue like a cyber-attack, natural disaster, or hardware failure. So how do businesses prepare for unexpected downtime to prevent a Facebook situation from happening? The answer is server virtualization (we'll there really could be many answers).

      In its reports, Facebook didn't disclose details about the incident, so we'll speculate. Had Facebook's downed server been a virtual server, creating a replica of that virtual server and standing it up rapidly would be the answer. If the server was physical, then virtualizing that server to build in redundancy could've been the answer. Here's more on server virutalization. To be fair, these are large generalizations and without knowing more details, the solution remains murky.

      Backing up or replicating your business’ data to the cloud builds in geographic redundancy to reduce the likelihood that an outage in one data center will impact your data in another data center. We advocate for a copy of data to be moved off-site and into the cloud at least 500 miles away from your business’ site. Additionally, the cloud allows your business to create an additional copy of your data, a protected copy that can be called upon amidst unplanned downtime.

      The Bottom Line 

      Facebook showed us that unexpected downtime does happen, even to the biggest companies. But enterprise business and small businesses alike are at risk. That’s why deploying a cloud solution to protect your data and systems can be the difference between successfully overcoming unplanned downtime and going out of business. Get in touch with a NewCloud Solutions Architect today to learn about cloud solutions customized to your business’ needs. 

      Topics: Cloud Computing, Desktop as a Service