If it seems like error messages have been popping up a lot more frequently on social-media sites, you’re not imagining things. Our favorite outlets for procrastination really do appear to be going through a rough patch, and disgruntled internet users have been leapfrogging between platforms to complain.
Now we have some evidence to back up these hunches. According to data from the last 36 months that Ookla’s Downdetector provided to Slate, it appears that outage times for Facebook and its subsidiary platform Instagram did spike in June and July. (Downdetector is a site that provides real-time updates of outages based on user reports.) Instagram, notably, had a major outage for about 12 hours from July 16 to July 17.
(The Y-axes of these graphs are in units of relative downtime according to an index calculated by Downdetector, which does not disclose absolute down times for “competitive reasons.” Downdetector also considers each platform separately for trend analyses.)
Data for WhatsApp, Facebook’s encrypted messaging platform, also indicates that outages increased to a lesser extent in July, though there were similar spikes in May, November, and December of 2017.
A spokesperson for Facebook, who stressed that the company is not in a position to comment on third-party data, provided Slate with a statement: “We strive to give the 2.7 billion people that use our products and services the best experience possible across our family of apps. At times, we experience temporary disruption in service. We work quickly to identify the issue and fix the problem at hand and restore service.”
Downdetector data also indicates that users were reporting relatively higher outage times on Twitter and Snapchat over the course of the last two months. However, Twitter saw comparable spikes in 2017, and Snapchat’s downtimes look to have been largely consistent save for a major outage in June.
It’s important to note that Downdetector sources its outage data from user reports, rather than directly from the platforms themselves, so it can at best give us a window into a problem, not a definitive measurement of one. Also, outages obviously vary in terms of the number and locations of users impacted, which can be difficult to discern from data.
While it’s hard to tell what exactly has caused these particularly downtime spikes, platform outages often occur in clusters because many sites rely on the same integral web services. Small problems in those services can snowball and lead mass site outages. For example, a botched software deployment by the content delivery network Cloudflare in early July took out thousands of websites, including Dropbox and Shopify. In June, a routing issue at Verizon also impacted major sites like Amazon, Google, and Reddit.
While these outages are largely a benign nuisance, they do remind us that the some of the biggest websites in the world rely on the much of the same infrastructure. When one piece topples, many more can follow.
