It’s a testament to how important Facebook has become in the web ecosystem that the social network’s performance issues yesterday didn’t just affect the site itself (and its 900 million users) but also a wide variety of other sites as well. Performance monitoring company Compuware APM, which analyses the performance of thousands of top sites, just sent us some interesting data about how Facebook’s problems yesterday correlated with significant slowdowns across major U.S. media and retail sites.
As our friends over at GigaOm pointed out today, “Facebook’s faltering didn’t lead to any noticeable traffic dip.” According to Compuware’s data, however, it did affect sites in other ways because of how tightly many media and retail sites integrate with services like Facebook’s “like” button, which was also affected by yesterday’s outage.
According to our own Josh Constine, yesterday’s outage wasn’t caused by anything external. Earlier reports indicated that Facebook was the victim of a denial-of-service attack, but a Facebook spokesman confirmed that the outage was not the result of third-party involvement. The outage started around 4:30pm yesterday and the site continued to struggle until around 3am PT Friday.
Overall, major Facebook outages are pretty unusual. When the site does go down, though, the ripple effects of these outages now affect large chunks of the web.