Microsoft's Azure virtual machine, cloud services down for many

Microsoft's Azure virtual machine, cloud services down for many

Summary: Over the past few days, Azure and other Microsoft cloud services have been down for a number of users worldwide.

TOPICS: Cloud, Microsoft

On August 18, starting just before 2 p.m. ET/11 a.m. PT, users across the globe began reporting problems with Microsoft's Azure virtual machines, web sites and other cloud services.


"Starting at 18 Aug 2014 17:49 UTC we are experiencing an interruption to Cloud Services and Virtual Machines in multiple regions," noted Microsoft officials on the Azure staus page.

As shown in the status-page screen caputre above, Virtual Machines is the service most affected by the outage, with web sites down in several US regions and cloud services down in other parts of the US.

"Virtual Machines, Cloud Services, Websites — Multiple Regions — Full Service Interruption," read the Microsoft-provided description of the issue.

I've asked Microsoft for an update as to what's going on and when the problems should be resolved. No word back so far.

Update (3:30 pm ET/12:30 pm PT): More Azure services, including Backup, Service Bus and Site Recovery —  are experiencing problems. An updated note on the Azure status page says:

"Starting at 18 Aug 2014 17:49 UTC we are experiencing an interruption to Azure Services, may include Cloud Services, Virtual Machines Websites, Automation, Service Bus, Backup, Site Recovery and possible other Azure Services in multiple regions. We are currently evaluating options to restore service."

Update (3:40 pm ET/12:40 pm PT): Azure HDInsight, Azure Mobile Services also are experiencing full service interruption in multiple geographies.

Update (3:45 pm ET/12:45 pm PT): Here's Microsoft's statement, via a company spokesperson:

“Per our message to customers on the Azure Status page, we are aware of an interruption with Azure services, including Virtual Machines, Cloud Services and Web sites, and are working with our engineering teams to resolve the issue as quickly as possible.”

Today's Azure service outages/problems comes just a few days after a major worldwide outage for Microsoft's Visual Studio Online service. Though Visual Studio Online runs on Azure, Microsoft officials attributed that outage to Visual Studio Online bugs, rather than Azure problems.

Last week, Azure was hit by a full service interruption in the Japan East region, management portal log-in problems and performance degradation in multiple Azure regions — all on August 15.

Microsoft CRM Online users were hit with problems of their own on August 15. Some users were down for 14 hours or more because of an unspecified issue which led to "some CRM Online organizations (being) disabled unexpectedly."

As far as I know, Microsoft's CRM Online service is still not hosted on Azure. But Microsoft management is pushing field sales and partners to make selling CRM Online a top priority in fiscal 2015.

Update (August 18 10 pm ET/7 pm PT): Looks like almost all is well now. Not sure exactly when the Azure  issues mentioned above were resolved.

Topics: Cloud, Microsoft


Mary Jo has covered the tech industry for 30 years for a variety of publications and Web sites, and is a frequent guest on radio, TV and podcasts, speaking about all things Microsoft-related. She is the author of Microsoft 2.0: How Microsoft plans to stay relevant in the post-Gates era (John Wiley & Sons, 2008).

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • All services working for me.

    EU datacentre, no issues.
    • Azure isn't needed

      to make hamburgers.
      • Really?

        Food manufacturing is cloud connected... I understand you flip burgers for a living... but what you didn't realize is that the little monitor above your head is connected to the cloud... its tells you how many burgers to flip....
    • No

      RemoteApp aka Project Mohoro is not running in europe.
  • Greatest thing since sliced bread

    And once those with common sense finally give up their struggle against the cloud and submit, our society will be completely at the mercy of hackers, terrorists, and rogue government agencies. If this is the reliability we can expect when things are "good", I can't imagine someone purposely trying to sabotage a telecom facility.

    Progress is wonderful.
    • Ive been arguing against the cloud since the first day I heard of the thing

      And what did I get back in my face:

      "But it can make collaboration so much nicer!!"

      B.S. Who cares. There can be perfectly workable software solutions that could create a collaboration environment or a single privatized service only for that if its so damn cool. But general cloud computing just puts the whole whack in someone elses hands. With so much cheap storage available, getting faster and cheaper all the time, its patently ridiculous to expect to convince a thinking person cloud based apps, storage and content should all be streamed all the time and nobody needs or should want their own copy of anything or find themselves needing to access data or use onboard applications. And that's what they are shooting for in the long term.

      We get nothing, they have everything and our computer at that point is nothing more then a rented key to unlock whatever cloud service we subscribe to to get at it all.

      That's the end game for the big IT -players. That's what they want, to be able to charge charge charge every month just to use what we used to own.

      Im not buying iit until there is nothing else left.
  • 512K problem??

    • yes

      according to loverock
    • good question

      That is what I was wondering as well. A big ISP here in Australia also lost the internet yesterday to (as of last night still hadn't found it again - but they were looking under chairs and everything) - so I wonder just how much of it is all caused by the 512K issue? I've run into problems intermittently accessing this site and a bunch of others over the past week.
  • Microsoft are so LOST.....

    They cant seem to get anyting right....
    WIN 8/8.1/8.1.update

    And they expect us to believe WINdows 9 (Threshold) will be any better.....

    All I can say Microsoft START listening to your 1million + clients out there and NOT your fanboyz, Own team, and give us paying clents what we want.

    WIN7 is great.
    WIN8/8.1/8.1.update is Cra#.
    Your Cloud is NOT stable enough for us to warrant paying for it.

    I look forward to your percieved offer of a free update from WIN8 to WIN9 in development for us to check it out.
    • Clown post

      That's why Microsoft's shares are up 12% over the past 12 months.

      Just another hater. Now go make me a sammich.
    • Ya, talk sensible or dont say anything.

      Blame it on MS, ya, that's the right solution. NOT.

      Complete fail on your part. If people listen to your jerky boy nonsense all they might do is buy into some other cloud service..

      They are all a garbage idea. MS themselves are no more the problem then the Atlantic ocean presents a problem for people breathing water like a fish. All bodies of water prevent breathing water like a fish and all cloud service ideas and models for the long term are solutions to a problem that only exists for big IT that wants a more steady stream of income.

      Drop the hate and just get real.
  • Monoculture

    It's really not fair to beat up on Microsoft. Both AWS and Rackspace which are of similar scale have had major meltdowns in the last few years. The problem is monoculture: these companies build cloud systems that are highly interconnected and vast scale, which ensures that any latent bugs or design problems can easily cause cascade failures, meaning that at least one full datacenter (and in the case of AWS, more than one) can be brought down when one of the bugs is triggered. Ultimately, the solution is clouds built on less interconnected, smaller clusters which are more failure-tolerant, especially if your application is deployed across more than one cluster. Smaller cloud providers generally provide this architecture because they can't afford the largest datacenters, are more likely to use VMWare's enhanced failover capabilities, and are less concerned about shaving fractions of a penny off costs through economies of scale because their value is in better services and reliability. The downside is that they don't or can't offer some of the cross-cluster services like multisite databases, but the increase in reliability is well worth the slight additional effort to set those up yourself.
    • Re: It's really not fair to beat up on Microsoft.

      Well maybe this is because Microsoft likes to bully and beat up on everyone else. The MS attitude certainly has created a lot of resentment with competitors and consumers alike.
      • Microsoft likes to beat up on others??

        Ha! Well! Surprise surprise!!! Oh my god, shock of shocks. amazement of all amazements!!

        I am really tired of all this Microsoft is a bad boy nonsense. Ive never heard of a corporation in my life that anyone has heard of that hasn't done outlandish things, often so far and above worse then the very worst of anything that Microsoft has EVER done.

        Do I need to start posting links to the well known "bean counter episodes" of a number of companies, perhaps some of the auto manufacturers are the best examples where they simply sat down and figured out the best dollars and cents net loss when comparing failed automobile functions that could result in death as opposed to doing complete recalls to fix the problem??

        Be serious. Ive read case after case after case, many right here on ZDNet about Apple, Google, Samsung, Intel, Facebook and other big IT names that have done all sorts of ridiculous things that have included costly law suites against opponents and unfair practices and so on that were designed as nothing more than doing their best to either undermine the competition or exploit some potential weakness to try and cripple some part of their market.

        All this Microsoft is "awful" nonsense carry's no more weight then pointing fingers at any major IT player and numerous other corporations claiming they are scum.

        Havnt you read long standing reports, pretty much undisputed that there is a real problem with thinking of corporations like a legal person because if they were such a person with a personality with interests anything like a corporation has, they would be psychotic without empathy for humans??

        This is not exclusive or even something close to being exclusive for a company like Microsoft. Other companies have spread some pretty thick lies to make themselves sound like the "first warm and fuzzy corporation", but its blatantly false because every single one of the SOB's immediately turn their backs on any thing that they are afraid will cost them something if there is a way around it, or out of it, even if it would be the more fair "human" way to do it.

        Drop all this pointless crap about Microsoft being so bad. Nobody is trying to hang any big auto makers, and here is Microsoft who dosnt kill anyone after the beans are counted yet we have the likes of pointless comments from you that "Microsoft likes to bully and beat up on everyone else".

        Ya, sure, they absolutely do, as all corporations do every single time they want to prevent losing a dime or need to make sure they make their dime. Show me something that makes MS worse than the rest and then you got a topic.

        And by the way, look if you like, but your not going to find something that does make them worse. They are all pretty much the same at the end of the day.


        Not real people.
      • That's relevant

        So it's not really about the stability of Azure, it's about your resentment against MS?
    • Monoculture

      You are right that monoculture plays a big role in allowing cascade failures. The Azure outage a few years back that was caused by a leap year (really?!!!) bug was made all the worse by Azure's automated patching and deployment. As it tried vainly to spin up more new VM's to handle the failures, those VM's were bit by the same bug and the outage spread.
  • hoisted

    Maybe they got hoisted by their own update.
  • The "Cloud" is not in your hands, and therefore not under your control!

    The farther away away you place your data, algorithms, and computing resources, the farther it is from your control. It may be one thing to lose your customers due to an outage, BUT HOW DO YOU EXPLAIN THE THEFT OF YOUR CUSTOMER/EMPLOYEE DATA?

    "It was the other guy's fault!?"

    Part of the problem is too much consolidation and movement towards globalization. The bigger the target, the more chance it will be attacked, i.e. MS operating systems and its apps.

    The sooner we stop squeezing the last penny out of everything we do, the sooner we will return to the reliable, trusted, and personal service of the 'mom and pop' local store.

    I'm afraid it will take a massive outage or data-theft (or a series of them) before we realize that going smaller is better.

    POOF Goes the CLOUD.
    • ABSOLUTLY 100% correect.

      "The bigger the target, the more chance it will be attacked, i.e. MS operating systems and its apps."

      Its an inescapable fact.

      For example, if you are a relatively small to mid sized company and have a real good security plan in place you can pretty much prevent casual hacking. Unless someone with some good skills finds a need to break into your otherwise small time outfit, your going to be safe from potential attack because half backed amateur random attacks will be unsuccessful against a company that's taken some relatively good security measures.

      Now sure, a big massive company like Apple of Microsoft or Google have some comparatively phenomenal security measures in place. They brush aside random amateur attacks like they were specks of dust, higher level attacks barely raise an eyebrow, aside form perhaps a comment like "notice that? I think it might have been some lame attempt at a break in".

      It takes a dedicated soul with superior skills and often resources to crack open one of the big players. But it happens, and we know it. We have seen admitted public acknowledgment of such many times over the years.

      Make a target large enough with enough potential information of value, or even just the potential to create high level chaos and it will attract the White Sharks. And given enough time and resources at their disposal, it soon becomes apparent that the risk of having your bits and pieces in the same vault as everyone elses gold may just mean you get your stuff blown to bits when the break in happens.

      There is nothing yet known of that can keep the right hackers out permanently when they have enough time and resources to create enough trouble for you. And the thing is, there are most definitely not enough of them to come knocking on everones gateway, so the big big targets get their attention.

      For a great many of us, security by obscurity works wonders so long as you cover your bases appropriately for your internet foot print.