Mark Zuckerberg does not seem to be interested in the comments that people make on his Facebook pages.

However, even if he could, it would be a long 145-day process, not including sleep, for him to go through the flood of comments he received after he apologized for the service meltdown last week.

After six hours in which Facebook, WhatsApp, and Instagram were offline, the Facebook chief executive and founder posted: “Sorry about the disruption today.”

Facebook blamed the disruption on a routine maintenance task – engineers at Facebook had unintentionally disconnected data centres of the larger internet.

An estimated 827,000 individuals responded to the apology of Mr. Zuckerberg.

There were many messages, from humorous: “It’s terrible. I had to speak to my family,” said one Italian user. Another was confused when someone wrote from Namibia, “I took mine to the repair shop thinking that it was dead.”

The very angry and upset: You can’t have all your doors closed at once. One Nigerian businessman stated that the impact was unprecedented. A second Indian businessman asked for compensation due to the interruption in their business.

If it was not obvious enough, what is becoming clear is how dependent billions of people are on these services. Not only is this for entertainment, but also essential communications and trading.

It is clear that this is not a rare situation. Experts suggest widespread outages becoming increasingly frequent and disruptive.

Luke Deryckx is Chief Technical Officer of Down Detector. “One thing that we have seen over the past several years has been an increased reliance upon a few networks and companies in order to deliver large amounts of Internet content.”

He says, “When one or more of these has a problem it affects not only them but also hundreds of thousands of services.” Facebook is used for example to log-in to various services and devices such as smart TVs.

According to Mr Deryckx, these internet “snow days” are happening now. “Something happens” [and]We all look at one another like, “Well, what are you going to do?”

His team from Down Detector, led by Mr Deryckx, monitor websites and web services for potential disruption. His belief is that large-scale outages of major services are increasing in frequency and severity.

Facebook is able to create huge impacts on not only the economy but also society when it has a problem. There are millions of users, possibly hundreds of billions of them, who just wait to see if a Californian team can fix it. This is an intriguing phenomenon that has increased in recent years.

  • October 2021: Facebook and Instagram were down for six hours due to a configuration error. Due to the flood of users visiting their sites, Twitter was also affected.
  • July 2021: More than 48 services including Expedia, Home Depot and Salesforce went down for about an hour due to a problem with Akamai’s Domain Name System (DNS). This follows an earlier similar incident at Akamai a month before.
  • June 20,21: Amazon and Reddit were offline for approximately an hour due to a bug that was accidentally triggered by Fastly, a cloud computing provider.
  • December 2020: Gmail and YouTube were all down for about 90 minutes simultaneously after Google announced that it had encountered an “internal storage limit issue”.
  • November 2020: An issue with Amazon Web Service’s Virginia facility, USA, caused thousands to lose access to third-party services, most of them in North America, for many hours.
  • March 2019, Facebook and WhatsApp went down, or had to be severely disabled for 14 hours following a “server modification”. Tinder, Spotify and other websites that login to Facebook were also affected.

Invariably, there will be a time when people fear that the disruptions are the result of a cyber-attack.

Experts believe it to be more common than you think. They say that the complexity of the internet’s systems, which are outdated and inefficient, makes this a complicated case.

Experts joked that the reasons behind Facebook’s outage were “older than Spice Girls” and had been “designed on the backs of napkins”.

Business Technology: More

  • Technology could detect disasters and prevent them from happening
  • Get energy from the fastest moving things on earth
  • Permanent chemicals must be replaced
  • Shipping crisis: Why large ships aren’t enough
  • Is it really that difficult to recycle a large jet?

Bill Buchanan is an Internet scientist. He agrees: “The internet’s not the large-scale distributed network DARPA (the Defense Advanced Research Projects Agency), tried to make. This could withstand any nuclear-strike.

It uses protocols that are essentially the same as those used when it connected from dumb terminals to mainframe computers. It is possible for the entire system to crash into the ground if it has a single problem in its core infrastructure.

Professor Buchanan suggests that improvements could be made in order to increase the resilience of the internet, but many of its fundamental elements will not change.

He said, “In general, systems work. You can’t switch certain protocols on the internet off for a day to make them new.”

Professor Buchanan suggests that we do not need to rebuild the structure or systems of the internet. Instead, we must improve its use to store and transmit data. Otherwise, there will likely be more outages.

The internet is too centralised according to him. Too much information is coming from one source. He explains that systems with multiple nodes can reverse this trend so no single failure can prevent a service’s operation.

There’s always a silver-lining. While major internet outages may affect the lives of users and their businesses, they also have the potential to help increase the resilience and availability of web services.

For example, Forbes estimates that Facebook lost $66m (£48.5m), during the six-hour outage, from the suspension, or exodus, of advertisers on the site. Senior executives will be more focused on stopping this type of loss from ever happening again.

Mr Deryckx stated that “They lost huge amounts of money in the day”, not only in terms of their stock price, but also their operating revenues.

You can also see that outages due to content delivery networks, such as Fastly and Cloudflare caused a lot of customer loss. This is why I believe these operators are trying to maintain online operations.


Share Your Comment Below



Please enter your comment!
Please enter your name here