Social media services Facebook, WhatsApp and Instagram appear to be recovering after an outage that lasted almost six hours.
All three services are owned by Facebook and could not be accessed over the web or on smartphone apps.
Downdetector, which tracks outages, logged tens of thousands of outage reports for all three platforms around the world.
The last time Facebook had a disruption of this magnitude was in 2019.
During the outage, Facebook’s chief technology officer Mike Schroepfer tweeted his “sincere apologies” to those affected.
Some people also reported problems using Facebook’s virtual reality headset platform, Oculus, and apps which require Facebook logins were affected, including Pokémon Go.
An outage of this scale for such a long time is rare. The disruption in 2019 left Facebook and its other apps mostly inaccessible across the world for more than 14 hours.
As yet there has been no official reason given for the problem. But online network experts speculated it may involve an error with DNS, or the domain name system, for Facebook sites.
DNS is often compared to an address book or phone book for the internet, pointing web browsers to the computer system which serves the website they are looking for.
Previous issues with DNS led to widespread outages of several major sites earlier this year.
In one of those instances, it emerged that the blackout was caused by a single customer of a widely-used service who changed their settings, triggering a software bug affecting a huge number of websites.
Several other tech companies, including Reddit and Twitter, poked fun at the social media giant’s predicament – prompting responses from the affected apps.
The disruption comes the day after an interview with a former Facebook employee who leaked documents about the company.
Frances Haugen told CBS news on Sunday that the company had prioritised “growth over safety”.
On Tuesday she will testify before a Senate subcommittee in a hearing titled “Protecting Kids Online”, about the company’s research into Instagram’s effect on the mental health of young users.
Outages for major websites aren’t uncommon. There was one last week on Slack for example.
What makes this so noteworthy is the size, scale and context.
Many outages that involve a domain name system usually get resolved fairly quickly. They are often localised too, with some people unable to open a website that can be viewed in another country.
This outage, however, was global, and affected all of Facebook’s many spin-offs.
The length of time it was off grid is also unusual. There were reports of “mayhem” in Facebook headquarters, as technicians scrambled to fix the problem.
What is also unclear is what has caused this. Generally outages are not caused by hacks, but that can’t be ruled out.
The week had already started off badly – after the whistleblower in the “Facebook Files” revelations revealed herself on Sunday.
But a bad week is fast becoming a terrible one for the social network.