If you’ve been using the Internet at all today then there’s a good chance you already know that Facebook, Instagram, WhatsApp, Messenger, and Oculus VR are all facing huge outages, as reported by The Verge. But now it looks like those sites and platforms are getting company. Gmail, Twitter, TikTok, Linkedin, Pokemon Go, Tinder, and more are all experiencing significant outages.
A quick trip to Down Detector shows similar outage spikes in what appears to be most of the services the website tracks. Gmail, Snapchat, TikTok, Discord, Zoom, Straight Talk, Cloudflare, Candy Crush, TMobile, Cricket Wireless, US Cellular, Amazon, Reddit, and many more all show outage spikes, with most of the spikes starting between noon and 3 pm Eastern time.
So what’s causing all these outages? Are we under attack? Is a bandwidth eating monster at the bottom of the ocean eating all the cables? Or are the end-times year, like at least one Twitter user jokes?
Well, from what we can tell, the main reason Gmail, Amazon, TikTok, and everything under the sun connected to the Internet seems to be going down has nothing to do with the apocalypse or any undersea creatures. The main problem, according to those we’re hearing from so far, is a DNS issue related to the initial Facebook outage. John Graham-Cumming — Cloudflare’s chief technology officer — explained it this way on Twitter.
So, DNS services are overtaxed dealing with the Facebook outages, and that overtaxing causes other outages, forcing Gmail, Twitter, Linkedin, and other services down.
One Reddit user claiming to be part of the Facebook recovery team had their post copied and pasted to Twitter, explaining not only what happened on the technical side of things, but how the human element has contributed to Facebook, Gmail, and other sites to go down. They explain that the Facebook outage is a symptom of Facebook’s peering routers going down. They also say the people attempting to get “physical access” do not have the sufficient knowledge to fix the relevant issues. The user goes on to say that another contributing factor is short-staffing in data centers due to the pandemic.