For every user looking to scroll the latest trends on Instagram, chat with their friends on WhatsApp or engage in Facebook, all these three services continue to experience an enormous outage and all the services remain offline.

Three of them are starting to come back online right after the BGP routing issues cause an over five-hour worldwide outage. Recently, at approximately 11:50 AM EST, all three platforms were suddenly unreachable, with browsers showing DNS errors when trying to open them.

While trying to connect directly to the following Facebook DNS servers, it was also impossible to reach them. As the DNS protocols tell your operating system and browser the IP address of a website, with the DNS servers unreachable, it was impossible to connect to them.

All about Recent Outage!

• At sharp 11.39 a.m. ET, Facebook, Instagram, Messenger, and Oculus VR went down. The outage is now turning off at the seven-hour mark.  
• Facebook has not given any reason for the outage but has tweeted that they are working on getting the issue resolved. 
• As of 6:10 p.m. ET, Facebook, Instagram, and WhatsApp are back online. We are still waiting for an official statement from Facebook regarding the outage.

BGP Routing Problem Outages

While at first, the concern arrived to be DNS related, it was later learned that the issue was far worse than that. As explained by Giorgio Bonfiglio, a Principal TAM at Amazon AWS, different Facebook routing prefixes had suddenly disappeared from the Internet’s BGP routing, effectively making it impossible to connect to any service hosted on their IP address.

BGP or the Border Gateway Protocol makes the modern-day Internet work and how a computer on one side of the world can connect to a device on the other. To make it easier to understand, the BGP routing protocol is the same as the Internet “postal system,” providing traffic from one (sovereign) system of networks to another.

While a network needs to be seen on the Internet, they are required to promoting their routes, or prefixes, with the rest of the world. If those prefixes are removed, no one else on the Internet has known how to connect to their servers.

As Facebook configured their organization to use a domain registrar and DNS servers hosted on their own routing prefix, when those prefixes were removed, no one could connect to those IP addresses and the services running on top of them.        

Initiating at 5 PM EST, the Facebook routing prefixes start to be seen on the BGP routing table at other networks. With these prefixes now being advertised on the Internet, users could connect to Facebook, Instagram, and WhatsApp once again.

It is unclear what caused the recent outage, but it was likely due to a configuration error, like many other BGP-related outages in the past. Our security researcher has reached out to Facebook to learn more about today’s outage but has not heard back at this time. Our earlier emails bounced as their emails servers were unreachable as well due to the outage.