Gareth Jones

For nearly 30 minutes in the early hours of December 3rd, anyone visiting Microsoft’s search engine was greeted by an error message.

A Microsoft statement said the outage was caused by a “configuration change” made to the site during testing. This, it said, had “unfortunate and unintended consequences” which included making the site unavailable.

Detecting the problem and fixing it took about 30 minutes, he said.

First question – why are people testing changes on a LIVE site? Doesn’t the largest company on earth have a test server somewhere?

Second question – why did it take 30 minutes to check the effect that this change had had on the site? When you change some of your code, don’t you test it immediately?

Final question – Where I work, if Google goes down someone notices pretty much immediately and yet no-one at Microsoft noticed Bing being down for 30 minutes? I guess a possible explanation for this is that they all use Google over there, but even then the developers responsible for this change should have been able to ignore Google for a second and test their changes on Bing before the support calls started coming in!