Facebook, WhatsApp, and Instagram are gradually returning. Why did they vanish? Except when it's BGP, DNS is always used.
Facebook's prolonged outage of a day is the longest and most severe in recent years. Facebook Messenger, WhatsApp, Instagram, and Instagram disappeared from the internet at around 9 AM PDT (on the U.S. West Coast), where the social media giant is located.
Market close was followed by a stock market plunge of around 5%, the company's stock falling around 5% from its Monday opening price. Services began to resume around midafternoon after Facebook sent a team to Santa Clara to reset the servers manually.
What makes this outage so unique is how offline Facebook was.
Facebook issued a short tweet in the morning to express its regret that some users were having difficulty accessing their apps and products. According to employees, they were unable access their offices. Staff called it a snowday because they could not get work done due to the outage that also affected their internal collaboration apps.
Facebook has not commented on the cause, but security experts stated that evidence suggested that there was a problem within the company's network. This caused Facebook to be disconnected from the wider internet as well as itself.
According to John Graham-Cumming (CTO of Cloudflare), the first signs of trouble occurred at 8:50 AM PDT in California. He said that Facebook disappeared from the internet in an explosion of BGP updates in less than two minutes. This refers to BGP or Border Gateway Protocol. It is the protocol used by networks to determine the fastest way to transmit data over the internet to another network.
These updates focused on BGP route withdrawals. Facebook sent an email to the internet stating that it was closing for business. It was like closing its drawbridge. Facebook was effectively isolated from the rest the internet because it had no routes to the network. Because of the structure of Facebook's network, route withdrawals also took out WhatsApp Messenger, Instagram Messenger, Facebook Messenger, and all other digital content within its digital walls.
Users began to notice problems a few minutes after BGP routes were removed. Rob Graham, the founder of Errata Security, stated in a Twitter thread that internet traffic that should have been directed to Facebook got lost on an unreliable internet connection and went nowhere.
Users noticed that their Facebook apps were not loading properly and they had problems with DNS. This is another crucial part of the internet's operation. DNS converts human-readable URL addresses into machine-readable IP addresses, which allows you to locate a website on the internet. Apps and browsers wouldn't be able to access Facebook's servers without kicking back what appeared to be DNS errors.
It is not clear why the BGP routes were removed. BGP, which is a well-known protocol, can be misused and maligned to cause massive outages.
It is more likely that a Facebook configuration update went horribly wrong, and the failure spread throughout the internet. Reddit thread by a Facebook engineer describing a BGP configuration problem long before it became widely known.
The fix may seem simple but the process of recovery can take several hours due to the way the internet works. Although internet providers typically update their DNS records every few minutes, it can take several days for them to propagate fully.
We are sorry to the large community of people and businesses all over the globe who depend on us. Facebook sent a tweet around 3:30 p.m. EST. We have been hard at work to restore access to our apps, and we are pleased to announce that they are now back online. We appreciate your patience.
