Earlier today, Facebook and related applications suffered a major crash, with external websites for Facebook, WhatsApp and Instagram, in addition to the company’s internal servers, down.
This is the worst break for the technology giant since 2008, when a mistake pushed Facebook out for about a day, leaving 80 million users without service. Today, it has approximately 3 billion users, and the outage has already reached its third hour.
Since 2016 Facebook has used its own domain registrar, but there may no longer be records of what makes it difficult for them to try to return online.
Facebook may have downloaded its own BGP routes to block attackers
Just before noon EDT, Facebook’s DNS was reported broken by Ars Technica reporter Jim Salter, according to tweet. “So @facebook’s DNS is broken this morning … TL; DR: Google anycast DNS returns SERVFAIL for Facebook queries; a.ns.facebook.com queries expire directly.” It seems that all four DNS servers on Facebook did not show any response, with the time to wait for cached areas for larger providers such as Google, Cloudflare and Level3. There were also major issues with other major apps under Mark Zuckerberg’s umbrella, with Instagram and WhatsApp reported. This was confusing because Instagram does not use conventional Facebook DNS – which pointed to a deeper problem.
At one point, the entire Facebook network became inaccessible because all of its Border Gateway Protocol (BGP) routes were withdrawn, according to the thread from Ars Technica journalist, citing @Cloudflare’s explanation. BGP distribution errors are not uncommon. They can sometimes come from direct attacks on the system, but “they usually leave SOME pockets of the world to function.”
“FB COULD have deliberately downloaded these routes, [for example,] to limit the ability of attackers to have access to compromised systems, “read a follow-up tweet from Salter, speculating why this is happening.
This is breaking news about why Facebook crashed on Monday, so be sure to contact us for more updates.