Facebook has said that it is working to avert a repeat of the outage that shut down its major platform hosting Facebook, WhatsApp and Instagram all of which went down across a massive period of nearly 6 hours on Monday.
During the period users were not able to access any of the platforms, and, according to reports, Facebook staff was likewise unable to access the company’s internal network. Now, many media publications are reporting that Facebook is blaming the outage on a “faulty configuration change.
Sources say that the outages were caused by a networking issue surrounding a new update that broke how Facebook advertises where its servers are to the internet using a Border Gateway Protocol (BGP), effectively making the servers undiscoverable.
A BGP is a system that the internet uses to route your website’s traffic to where it needs to go as quickly as possible. Since there are probably untold thousands of internet service providers, backbone routers and servers from all over the world responsible for making sure user data makes it to sites like Facebook, WhatsApp and Instagram, there are innumerable routes the data can take. Some of these routes are shorter, some are much longer.
BGPs make sure data takes the most efficient routes. Oftentimes BGPs are described as air traffic controllers as they control masses of data and ensure the flow. The “faulty configuration change” meant that the BGP could not direct traffic to Facebook and its apps because it stopped telling the ISPs, routers and servers of users where the company’s data centres were.
Essentially, Facebook, Instagram and WhatsApp weren’t “down” in the strictest sense, it was just that no one could find them.