At 4:51 p.m. on Monday, Spanish peninsular time, the tools of the social corporation Facebook stopped working. From one second to another, its 3,000 million users stopped being able to send messages or consult the publications of their contacts. Companies that do not have alternative methods of reaching the public have lost their storefronts and customer service avenues. A technical failure, the company says, caused WhatsApp, Instagram and Facebook to turn off and left 60% of Internet users who used their services globally in the dark for six hours.
The author of the Facebook leaks: “Between society and profits, the company always opted for money”
“The underlying cause of this interruption has also affected many of the internal tools and systems that we use in our daily operations, which has complicated our attempts to quickly diagnose and resolve the problem,” Santosh Janardhan, chief, apologized in a statement. infrastructure manager, which has apologized on behalf of the company for the inconvenience.
Janardhan’s explanation confirms the first suspicions of industry professionals who pointed out that the historic drop was due to a problem with the BGP protocol (Border Gateway Protocol). This mechanism, belonging to one of the deepest parts of the digital infrastructure, works as a ‘direction map’ for the data. If the Internet is a network of networks, BGP is the instructions for connecting one network to another. Problems with Facebook’s BGP erased the path to their systems from that map. The machines, not knowing where they had to send the users’ information queries, did not send them.
“Our engineering teams have found that configuration changes to the backbone routers that coordinate network traffic between our data centers caused problems that disrupted this communication. This disruption of network traffic had a cascading effect on the way in which our data centers communicate, which caused the interruption of our services “, exposes the person in charge of infrastructure of Facebook.
Translation: For the machines, at 4:51 p.m. on Monday WhatsApp, Instagram and Facebook “disappeared from the Internet”, as explained by Cloudflare, one of the network infrastructure companies that was able to understand and explain what was happening faster.
Chronicle of a global blackout
The Cloudflare experts they wrote a chronology of the cascade of failures referred to by Santosh Janardhan. “BGP allows a network (say Facebook) to advertise its presence to other networks that make up the Internet. As we write, Facebook is not advertising its presence, ISPs [Proveedores de Servicios de Internet] and other networks cannot find the Facebook network and therefore it is not available, “they documented.
“The big routers that make the internet work have huge, constantly updated lists of possible routes that can be used to deliver each data packet to its final destinations. Without BGP, internet routers wouldn’t know what to do, and the internet wouldn’t work. “they continued. That is exactly what happened to Facebook.
How did it happen? This is where the social network has not gone into details. What Cloudflare’s systems registered is that at 5:40 p.m. (Spanish peninsular time) Facebook sent a large number of updates to its BGP. This is an anomalous event, since these routes are only changed sporadically. “We saw a spike in routing changes from Facebook. That’s where the problems started.”
Although the explanation of what happened is complicated, the consequence was very simple: “Facebook and its sites had effectively disconnected from the Internet,” the company concluded.
From Transparent Edge Services, a Spanish company that offers network infrastructure services, they came to the same conclusion. “Around 5:50 p.m. Spanish time, Facebook withdrew (for reasons still unknown) its routes from the BGP ad system, thus removing themselves from the Internet,” said Diego Suárez, the company’s Chief Technology Officer. “If, for whatever reason, Facebook decides one day to withdraw its business from the Internet, this is the path it should follow,” the expert continued.
We have no evidence that user data has been compromised as a result of this outage.
– Facebook Infrastructure Manager
Due to the technical complexity of the failure and the fact that all Facebook platforms committed it at the same time, specialists see it unlikely that it was a cyber attack. “We want to make it clear at this time that we believe that the main cause of this service interruption was a faulty configuration change. We also have no evidence that user data has been compromised as a result of this interruption,” they have stressed from Facebook .