Bing bonked by service outage Thursday, Microsoft configured the wrong server
For about 40 minutes starting at approximately 9:30 pm EST Thursday evening, by Betanews estimates, the main page of Microsoft's Bing Web site was inaccessible to users. In its place was a message filled with hexadecimal code, leading off with the message, "This isn't the page you wanted!"
No kidding. But was Bing being hijacked? As Microsoft acknowledged this morning, its own administrators were responsible for the outage.
"The cause of the outage was a configuration change during some internal testing that had unfortunate and unintended consequences," read a blog post from Online Services Division Senior VP Satya Nadella. "As soon as the issue was detected, the change was rolled back, which caused the site to return to normal behavior. Unfortunately the detection and rollback took about half an hour, and during that time users were unable to use bing.com."
Nadella's explanation seemed to imply that Bing doesn't have a backup server, or a "production server" and a "staging server." In a normal environment, configuration changes are made to one server offline, while the other one continues to face the public. Is Bing's public-facing Web server the only one there is, as Nadella implied?
Having experienced enough problems this week with Microsoft's having not said something being interpreted as Microsoft having said something, Betanews posed the question directly first to Microsoft itself. Yes, the company replied, Bing does have a staging server.
And that was actually the problem, as its spokesperson revealed to Betanews this morning. Bing's admins were configuring the wrong server by mistake.
"A configuration change was mistakenly propagated to production from staging," the spokesperson told us. "It was supposed to stay in the test environment -- it was a mistake. As we mentioned in the blog post last night, we are looking at process to ensure it does not happen again."