The Microsoft cloud went south for a couple of hours Thursday night. The silver lining is that Microsoft appears to be doing better at informing its users what's going on -- although a new third-party cloud tracking service beat Redmond to the punch.
At about 7:18 p.m. Pacific Time, people started hollering that they couldn't get into their Hotmail accounts. SkyDrive was down. So, too, were MSN, Azure, and Office 365. Again.
At 8:43 p.m., @Office365 tweeted, "We are investigating service interruption. Will update as soon as more details are available or within hour." At the same time, @MSN tweeted, "We're currently experiencing an outage at www.msn.com We're aware of the issue and working to resolve it. Thanks for your patience."
At 9:14 p.m., @hotmail tweeted: "We're aware of an outage impacting Hotmail, SkyDrive, and our other properties. We are actively working to resolve this issue. More to come."
At 9:26 p.m., the official Windows Live Status site reported: "We're aware of a problem with Hotmail that's affecting some people. We're investigating and will provide an update by Sept. 9 11:30 p.m.. Some people might be experiencing sign-in issues."
Anecdotally, it looks like customers in Asia-Pacific got hit first, then Europeans, and then Americans. Service came back up for different locations at different times, but somewhere between 10 p.m. and midnight Pacific time, everything was back to normal.
For technical reasons -- apparently related to DNS problems that may have triggered the whole hiss -- Office 365's Service Dashboard didn't work.
Microsoft's getting better at reporting outages, Service Dashboard or not. At least in this case, Twitter seems to be the official voice of choice.
But the real vox populi on this outage didn't come from Microsoft or even from Twitter. It came from downrightnow.com, a site that specializes in tracking online outages. Remarkably, downrightnow reported the outage at 7:18 p.m., more than an hour before Microsoft even hinted there was a problem.
Downrightnow uses reports from users made on their website to come up with their early-warning system -- cloudsourcing, if you will. It also relies on Twitter reports; for example, the hashtag #downrightnow #hotmail automatically notifies the site that Hotmail's headed south. The site also says it takes manufacturer's reports into account, but apparently it doesn't wait for official word.
Well, outages happen. Minutes before the Microsoft outage, the Apple Store came tumbling down -- an event that recurs with such frequency that there's a website devoted to tracking Store outages. A very cursory look shows extended Apple Store outages in April, May, July, and August. This one started about 7 p.m. Pacific time. (No, Apple didn't pull the store down to list new products, much less the new Mac Pro. It just curled up for a few hours.)
Google Docs went down worldwide for at least 30 minutes on Wednesday. Such outages aren't a question of whether, but when -- and how the vendors respond. In the case of the Google Docs outage, reporting was swift, accurate, and easily accessible on the Google Apps Status Dashboard.
In the case of the Office 365/Azure/Hotmail/SkyDrive/MSN outage, notification was sporadic, fragmented, and slow ... but a whole lot better than it used to be. Progress!
This article, "Blow-by-blow: Another Microsoft cloud outage," was originally published at InfoWorld.com. Get the first word on what the important tech news really means with the InfoWorld Tech Watch blog. For the latest business technology news, follow InfoWorld.com on Twitter.