The company said in a tweet: "Yesterday, as a result of a server configuration change, many people had trouble accessing our apps and services.
"We've now resolved the issues and our systems are recovering. We’re very sorry for the inconvenience and appreciate everyone’s patience."
Yesterday, as a result of a server configuration change, many people had trouble accessing our apps and services. We've now resolved the issues and our systems are recovering. We’re very sorry for the inconvenience and appreciate everyone’s patience.— Facebook (@facebook) March 14, 2019
One post on the North American Network Operators' Group mailing list said the issue was caused by Facebook pushing an update to its code that manages cookies, that had a rather severe bug in it that resulted in a large flood of requests to their database servers.
Facebook was not specific about the issue that caused the biggest outage the site has experienced, and a network professional has now done a backflip on the explanation he offered on Thursday that the outage was likely to be due to a border gateway protocol routing leak.
Roland Dobbins, principal engineer at application and network performance management products provider NETSCOUT, had said on Thursday: “At approximately 12:52PM EST on 13 March, it appears that an accidental BGP routing leak from a European ISP to a major transit ISP, which was then propagated onwards to some peers and/or downstreams of the transit ISP in question, resulted in perceptible disruption of access to some well-known Internet properties for a short interval."
But he later told Ars Technica reporter Dan Goodin that an "internal miscommunication" had led to PR people sending out the claim about BGP.
Roland Dobbins, principle engineer at Netscout's Assert team, says he has no data whatsoever to support the claim that a BGP leak is the cause of today's Facebook or Instagram outages. "There was an internal miscommunication here," he says of the email PR people sent to reporters— Dan Goodin (@dangoodin001) March 13, 2019
Jeroen Wunnink, who works for network service firm GTT, outlined a possible reason why a BGP leak had been advanced as the reason for the Facebook outage.
"The route-leak was something different that seems to have mainly hit west-Europe between 16:52 UTC to 17:08 UTC,." he said in a post to the NANOG mailing list.
POLICE HAVE HAD TO TELL PEOPLE TO STOP CALLING 000 ABOUT FB BEING DOWN. FFS PEOPLE, JUST BITCH ABOUT IT ON TWITTER LIKE A NORMAL PERSON #FACEBOOKDOWN— The NT News (@TheNTNews) March 14, 2019
"There’s a few people in the *NOG communities still digging at the complete details of that right now, but it currently points to have originated from AS200020, impacting a few large upstreams for a short period of time.
"So unless this leak caused a catastrophic cascade in FB’s network somehow, it seems to be unrelated. It looked like a valid suspect because timing was very similar between the start of the FB outage and the leak."