Why didn’t Facebook and Messegner work? Hundreds of thousands of reports

On Tuesday afternoon, Facebook, Messenger and Instagram users around the world experienced serious performance issues with these applications. Many of them were automatically logged out of the websites, which raised concerns that their accounts had been hacked. Some people tried to reset their passwords because of this, but this did not solve the problem. It turned out that there was a global failure of all websites belonging to the Meta group.

Reports on downdetector.pl began to flow en masse after 4 p.m. Their number was growing very quickly. At 4:17 p.m. there were over 76,000 reports, and by 4:30 p.m. the number was over 199,000. At the peak, there were already over 500,000 reports of failures, which is an unprecedented situation.

What’s going on with Messenger? Global blackout

Late Tuesday afternoon, Andy Stone, a spokesman for Meta, admitted that the company is aware of problems with access to its services. “We are working to solve the problem,” he added. The failure was brought under control after 5:30 p.m. Polish time. According to Meta itself, it was caused by a “technical glitch”.

There was a technical issue this morning that caused people to have difficulty accessing some of our services. We resolved it as quickly as possible and apologize for any inconvenience.

– said Andy Stone in an entry published on website X. However, Mark Zuckeberg’s company did not explain what exactly the defect was.

Elon Musk mocks Facebook. “If you are reading this, it means our servers are working”

Due to the outage of Facebook and Instagram, Elon Musk’s website X saw a significant increase in traffic on Tuesday. Many people were looking for information there about the causes of Meta’s service failure.

Elon Musk decided to use the opportunity to mock his rival Mark Zuckerberg. “If you are reading this entry, it means that our servers are working,” he wrote. He then posted a rather explicit meme with the famous penguins from Madagascar.

This wasn’t Facebook’s biggest failure

This isn’t Facebook’s first major outage. “, which lasted as long as six hours. Then the culprit turned out to be an incorrect update of the BGP protocol, which is the basis for the operation of the modern Internet. This system “controls” the traffic in the network and decides what route individual data packets should take in order to reach them as quickly as possible. place A to place B.

Why did it take Facebook employees so long to deal with the problem? It turns out that the reasons were quite bizarre. As Alex Hern, a technology journalist for the Guardian, explained in a series of entries on Twitter, the internal policy of the company, all of whose systems run on Facebook servers, turned out to be to blame.

Facebook accidentally sent out a deep-level routing protocol update that basically says: ‘Hey, we don’t have any servers anymore’ (…). Usually, such an error can be easily fixed, it would be enough to send another protocol update saying, in simple terms: ‘Don’t worry. However, we do have servers. They are here!’

– explains the journalist.

The problem is that sending such an update turned out to be impossible because, as Hern explains, “EVERYTHING in Facebook runs through Facebook.” As a result, it turned out to be impossible to repair the problem remotely because Facebook’s servers were cut off from the outside world. To deal with the failure, FB employees had to physically appear in the company’s server room and restart the servers.

Here another problem arose. Opening the door to the server room turned out to be impossible because the access cards… also work based on Facebook servers. The problem could have been solved by the security chief who had a physical key. However, something went wrong here again, because the alarm system that allows quick contact with the security chief – yes, you guessed it – runs on Facebook servers.