[ad_1]
Monday’s major Facebook outage has ignited the internet with worry and guesswork as to whether it was caused by an engineering error, a malicious attack, or even something else.
But, about an hour after the outage, which also saw Instagram and WhatsApp down for the five-hour count, a Reddit user named u / ramenporn, who claimed to be part of the “recovery team” for the issue. in progress, tried to explain what was going on.
Facebook technicians had to be in physical proximity to routers
Some early responses appeared to come from Reddit. Specifically, they appeared to come from people who weren’t allowed to speak.
“As many of you know, the DNS for FB services has been affected and that’s probably a symptom of the real problem, and that is, BGP peering with Facebook peering routers has gone down, very likely due to a configuration change that went into effect shortly before the outages occurred (started at around 1540 UTC), ”read the initial post on Reddit by an alleged Facebook insider.
He continued, “There are now people trying to access peering routers to implement fixes, but people with physical access are distinct from people who know how to actually authenticate to systems and devices. people who actually know what to do, so there is now a logistical challenge with the unification of all this knowledge. “
“This is also partly due to the reduction in staff in data centers due to pandemic measures,” added Reddit user u / ramenporn.
Shortly after the alleged insider began leaking potentially sensitive information about Facebook’s desperate attempts to regain control of its domains, the Reddit user’s post was deleted. But the Wayback Machine kept an archive of the information.
Another Reddit user seemed to think this ordeal would lend credibility to his case against Facebook management regarding why we need knowledgeable staff available in major data centers. Other users have complained about the challenge of fixing an issue like this when the network itself goes down.
“The problem is that when your network goes down, even if you enter through a backup DSL connection or something to the data center, you cannot switch from your hop host to something else,” it reads. an article by mike_d.
Reddit user u / ramenporn also said Monday’s outage was likely due to Facebook’s network engineers accidentally locking themselves out of the larger system during a configuration change.
Once that happened, that meant the only ones who could do anything had to be near the physical routers in Facebook’s data center, to bring back the servers. Fortunately, someone (or possibly several extremely stressed-out people) fixed the underlying issue, as Facebook and Instagram resumed service shortly before 6:00 p.m. EDT.
This story was in development and was regularly updated as new information became available.
[ad_2]
Source link