Facebook said on Friday that users around the world had trouble accessing its services for several hours due to a change in its setting, just days after a major crash in a similar fashion.
“We sincerely apologize to anyone who has not been able to access our products in the last two hours,” a Facebook spokesman told AFP at 21:30 GMT.
“We fixed the problem and now everything should be back to normal.”
Down Detector Internet Problem Tracker launches three hours before reports of problems with accessing or using Facebook and its photo-centric Instagram network and Messenger and WhatsApp.
Facebook attributed the problem to the configuration change on its computer site and said it affected users of the social network and Instagram, Messenger and the workplace worldwide.
People flocked to Twitter to express their frustration.
“What’s on Instagram?” Read a tweet that included a picture of cartoon character Bart Simpson sitting in a corner in apparent punishment.
“It’s not even 4 days. It’s already down again.”
“Problems again on Instagram, Facebook, Facebook Messenger and WhatsApp!” Read a lament in the Downtector chat forum.
Hundreds of millions of people were unable to access Facebook, Instagram or WhatsApp on Monday, outlining the world’s reliance on Silicon Valley-owned sites.
In an apology blog post, Santosh Janardhan, Facebook’s vice president of infrastructure, said the day’s crash was caused by “configuration changes” in routers that coordinate network traffic between data centers.
The problem with cyber experts is called the BGP, or Border Gateway Protocol – a system that selects the fastest way to move Internet information packets.
Sami Slim of Telehouse, a data center company, likened PGP to an “Internet parallel to air traffic control.”
In the same way that air traffic controllers sometimes make changes to flight schedules, “Facebook updated these routes,” Slim said.
But there is a significant bug in this update.
How or why is not yet clear, but a message was sent to the Internet based on Facebook’s routers, announcing that the company’s servers were no more.
According to Facebook, Friday’s crash was not related to the previous week.
Experts say Facebook’s technological infrastructure is unusually dependent on its own systems.
Social media crash is not uncommon: Internet builder Tooldestor says Instagram alone has experienced more than 80 in the United States over the past year.
Facebook’s services are important to many businesses around the world, and Facebook accounts are commonly used to sign in to other websites.
Facebook’s processors use my billions of people every month, which means that crashes will affect a large part of the world’s population.