A brief outage in Google's services last week nearly halved the web traffic recorded by analytics outfit GoSquared. The downtime reportedly lasted only a few minutes on Friday afternoon, but it was significant to affect most of the company's major products including Gmail, Calendar, Drive and YouTube. While the hiccup probably annoyed some folks, plenty used the opportunity to joke around:
GoSquared's graph (below) shows that Google fixed the problem relatively fast, but it still caused a 40% drop in the analytics firm's global pageviews. "That's huge," GoSquared's Simon Tabor wrote, also noting that the company also saw a spike in pageviews as services returned to normal activity and users began reaching their destination -- undoubtedly after testing the resilience of their F5 key.
The outage lasted five minutes or less depending on the service and source you look at, but Google said that 50-70% of requests coming into the search giant between 15:51 and 15:52 PT received errors. Services were largely restored within that one-minute window and they were entirely fixed in the following four minutes according to Google. GoSquared logged the downtime from 15:52 to 15:57 PT.
In somewhat related news, the pace at which Google handled its outage has left some users questioning the reliability of Microsoft's rival cloud offerings. Outlook, SkyDrive, Calendar and People experienced issues including sporadic outages that started last Wednesday. The problem began after a caching service for Exchange ActiveSync failed and things snowballed from there, according to Microsoft.
"We realize that we have a responsibility to the customers who use our services to communicate and share with the people they care most about, and we apologize for letting those customers down this week. Our first priority is to the health of the services, and we will learn from this incident and work to improve the experience of all our customers," Microsoft said. To help prevent future service interruptions, the company has allocated more network bandwidth to the affected infrastructure and it has improved the way error handling works for devices using Exchange ActiveSync.