Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Cloud Microsoft Software The Almighty Buck IT

Office 365: Suffer 18 Days' Outage, Still Pay Half Price 137

Posted by timothy from the you-ungrateful-whelps dept.
rtfa-troll writes "Microsoft is preparing its customers for plenty of outage time according to the Register, with a scheme for Office 365 which will give customers some money back. The offer seems to be Microsoft's answer to Google offering a '100% uptime guarantee' (they even pay for maintenance time) The most interesting thing about the scheme is that you can have a one and a half day outage every month (or is that 18 solid days a year?) and still expect to pay half price. I wonder Microsoft have put the Sidekick management in charge of their customer's data. Looking forward my expense forms have getting eaten by the cloud so I have to fill them in again."
This discussion has been archived. No new comments can be posted.

Office 365: Suffer 18 Days' Outage, Still Pay Half Price More

Office 365: Suffer 18 Days' Outage, Still Pay Half Price

Comments Filter:

  • That's awful, a plain-jane Windows server manages way better uptime than that!

    • Re: (Score:2)

      by jimicus ( 737525 )

      Until such time as a major component fails, that is. The point of cloud hosting is that the system is designed so that's a non-issue.

      (Of course, that assumes your provider has actually designed their systems to account for component failure. I can't count the number of companies that don't appear to have done so.)

      • Re: (Score:2, Insightful)

        by Anonymous Coward

        Then 4 hours later dell gets you the replacement part. Not 18 days later.

        • The 18 days comes from the product of 1.5 days/month * 12 months.

          I think it's a bit high regardless (1.5 days could easily fall on a weekend that I might need or even want to work), but it's misleading (at best) to suggest it's 18 days of straight offline time.

          Even worse, no where in the article does it actually state 1.5 days. Anywhere. I must be new here, but here is the relevant quote from the article:

          Under the service level agreement, customers receive 25 per cent off their monthly payment if uptime fal

          • It looks like someone did 365 days by about .9505 (still technically over 95% uptime, requiring half payment) and divided that by 12 to get 1.5 days per month.

            I guess it's a possible outcome... but clearly someone was looking for the worst possible way to represent the ToS. Maybe that's part of diligent research for your company... but certainly doesn't tell the whole story (like you did).

            But then... an overblown /. summary? Not surprising.
            • No; I did (pretty much exactly)

              rtfa-troll@ubuntu:~$ bc -l
              bc 1.11.28
              Copyright 1991-1994, 1997, 1998, 2000, 2004, 2006, 2011 Free Software Foundation, Inc.
              This is free software with ABSOLUTELY NO WARRANTY.
              For details type `warranty'.
              365 * 0.95
              346.75
              365 - 347
              18

              And yes, when dealing with Microsoft, I do try to work out what is the worst possible interpretation of their contract. It's not sufficient to understand how they will screw me, but it's a start. Anyway, you were pretty close.

          • Re: (Score:2)

            by praxis ( 19962 )

            The 18 days comes from the product of 1.5 days/month * 12 months.

            I think it's a bit high regardless (1.5 days could easily fall on a weekend that I might need or even want to work), but it's misleading (at best) to suggest it's 18 days of straight offline time.

            Even worse, no where in the article does it actually state 1.5 days. Anywhere. I must be new here, but here is the relevant quote from the article:

            Under the service level agreement, customers receive 25 per cent off their monthly payment if uptime falls below 99.9 per cent to 99 per cent, half of the sum back if it falls below 99 per cent and a complete refund for anything under 95 per cent.

            King said clearly Microsoft would prefer it had no issues but claimed: "the processes in place are robust and financially backed, if you look across cloud providers in the market that is unique."

            In other words, it's just like Google's service, only they don't claim 100% uptime, which is unlikely to be realistic (even Gmail has failed on numerous occasions). And, they pay you if it falls below 99.9% uptime. Considering that you still get the benefit of local deployment, as well as the cloud, I'd say that's actually a good deal.

            Not one I have any interest in paying for, but it sounds a lot better than Google's unlikely claim.

            You quote the part where it states 18 days for the year, or 3/2 days for the month. The authoer left the calculation as an exercise to the reader, but it's pretty clearly spelled out there.

            • Yes, it was left up to the reader to find the second worst possible scenario. The only thing worse would be anything less than 94.99%, which would mean that you get the entire service for free.

              Why not mention 365 possible days of downtime, or even 19 days per year? After all, it's just as possible as 18 days per year for 50% off.

              Not only was the summary intentionally misleading, but it was a downright lie when he mentioned Google's "100%" claim that I took at face value. The Google article [infoworld.com] never even mentio

          • Re: (Score:2)

            by rtb61 ( 674572 )

            Then of course you must add in your ISP's guaranteed uptime, what you don't have one. Well, what about all those hops between you and yours and your cloud, all those connections that must work for you and yours to work.

            The cloud always sounds nice but with all those cheap profit hunger bastards between you and your data, that cloud can often be end up being droughts and floods rather than a steady rain. Sounds better as a backup for internal primary services (use the cloud only in emergencies, concurrent

      • Well let's say this shitty Windows box boots from a single hard drive and it fails once per week (maybe someone ran scandisk on a drive with large files and the dumb thing filled up the RAM and then went into page file thrashing because that's what scandisk does with large files for some reason, even in Win7!!!).

        Let's say it takes 10 minutes to swap the drive and 2 hours to restore it from a backup. That's 2 hours and 10 minutes of downtime per week, or 8 hours and 40 minutes of downtime per month. Still le

    • Are we talking about a total outage or are we talking about a percentage of users left unable to get their data?

      I ask because I imagine it's a huge technical challenge to have constantly editable data synchronized across all the machines in the 'cloud'. (I will be up front and admit that I don't know a lot about how something like this works.)

      • The only problem you might have is two people trying to edit the same document at once, but you'll have that no matter what technology you're using. The storage on this should work like shared storage on a giant RAID5/6 array so that many servers have no problem accessing the same data and some hardware can go down without causing any trouble.

        • Well that's great if all the servers are in the same location. But what about cases where you've got some servers in LA and others in NYC, like Amazon and I think Google does, how do they sync up?

          • Re: (Score:2)

            by Tynin ( 634655 )

            Well that's great if all the servers are in the same location. But what about cases where you've got some servers in LA and others in NYC, like Amazon and I think Google does, how do they sync up?

            I'm sure their are numerous ways to approach this problem. But a lot of it is just having a huge pipe going between your datacenters to connect your environments. Then it is just a matter of having numerous near identical platforms/clusters that you can take in and out of your load balancers at each site.

            So say you had servers in two different geographical locations, and in both locations you had at least a few clusters. You keep at least one or two cluster hot(live customer facing) in both sites and sy

  • Typo in title (Score:3)

    by Bob the Super Hamste ( 1152367 ) on Thursday June 30, 2011 @03:14PM (#36626248) Homepage
    I don't think it is called Office 356.

    • 365-356=9=half of the maximum expected downtime. Makes sense.

      • Re: (Score:2, Insightful)

        by pushing-robot ( 1037830 )

        So it's advertised as Office 365, you're paying for Office 356, but getting Office 347.

        Yeah, that sounds like Microsoft.

    • Re: (Score:2)

      by Lifyre ( 960576 )

      I thought it was a clever attempt at humor... then I realized that timothy would have to be really bad at math for that...

    • Re: (Score:3)

      by Niris ( 1443675 )
      But you're not concerned with the grammatically incorrect 'I wonder Microsoft have put the Sidekick management in charge of their customer's data"?

    • that's only one of the myriad errors in grammar in that summary.

    • the month to month SLA part is entirely comedy though. If they average less than 99% uptime over a year, even if it's one particular bad month, people are going to drop this service instantly at that point. A consumer may be stupid enough to forget month to month, but a business won't.

      • You discount the money hole syndrome. I a business dig's itself into a hole, you the only way to fill it back up it to pour more money into it.

        Seriously though, there are a lot of executives happily looking forward to the day they get to fire that IT guy who keeps making him feel bad. I bet they could call it Office 256 and plenty of businesses would buy it.

        • Re: (Score:1)

          by aix tom ( 902140 )

          I wonder how many executives will curse the day they fired a "IT guy that made them feel bad" and replaced him with a "Cloud that ignores them" in case of problems.

          • Re: (Score:1)

            by N!k0N ( 883435 )
            none. They'll find a reason to curse the IT guy for allowing them to go to the cloud in the first place... regardless of the fact he was vehemently against said idea and was fired for saying as such.

          • Probably all of them... it will just make it that much more fun to watch.

            "Patty, get Tom on the line. My Word isn't working and I can't check my e-mail."
            "Sorry Mr. Jamison, you fired Tom and replaced him with a BORG"

  • Of outage? Looking closely at google apps...

    • Re: (Score:2, Informative)

      by Anonymous Coward

      Microsoft is promising to be twice as good as Google Apps. Seriously, check it out:

      http://www.google.com/apps/intl/en/terms/sla.html

      For a service level of:

      >99.9% - Microsoft: full price, Google: full price
      >99.0% - Microsoft: 25% off, Google: 10% off
      >95.0% - Microsoft: 50% off, Google: 25% off
      <95.0% - Microsoft: 100% off, Google: 50% off

    • Re: (Score:2)

      by kenh ( 9056 )

      Google Apps SLA is almost EXACTLY the same - 50% credit (15 days) for 95% uptime in a month - guess how long an outage you can have with 95% uptime? Just about a day and a half...

  • I thought it was supposed to be a joke, but 365-18 = 347, so I guess not. If the story was titled "Office 347," well, that'd be pretty witty.

    • That's because it was only half of the joke. The other half is that you get 635 days of up time on a two-year contract.

  • Data loss is your own fault... (Score:4, Insightful)

    by bjwest ( 14070 ) on Thursday June 30, 2011 @03:28PM (#36626454)

    Looking forward my expense forms have getting eaten by the cloud so I have to fill them in again.

    Especially this early in the life cycle of this "cloud" crap. Any expectation of not loosing your data if you don't keep a backup yourself is entirely your own fault.

    Besides, I though we left terminal computing (either smart or dumb) back in the '80's. Screw that crap, I'll keep my data and aps on my own computer, thank you.

    • Re: (Score:2)

      by Inda ( 580031 )
      Surely people can see the good too?

      I'm fed up while having to double my storage capacity every few years. I'm fed up with having dusty DVDR backups laying around and I'm fed up with burning them. It's time to let someone else worry about it all; I have better things to do.

      • Re: (Score:1)

        by Anonymous Coward

        You better get your data off those DVDRs bro. They don't last forever (and the cheap ones don't even last a few years). I learned that one the hard way.

      • Crashplan allows you to back up to a local HDD as well as to their data centers, and automatically, all you need is a connection and your computer to be turned on. It's a hell of a lot easier than managing and tracking DVDs.

    • Re: (Score:2)

      by MacTO ( 1161105 )

      If you want to keep your data and apps on your own computer, then you are free to do so. If someone else wants to keep their data and apps in the cloud, they have that option too. Or, better yet, pick and choose based upon the situation. After all, both standalone and cloud computing have their benefits and drawbacks.

    • "Any expectation of not losing your data if you don't keep a backup yourself is entirely your own fault."

      That applies anywhere.

    • Re: (Score:2)

      by sorak ( 246725 )

      Besides, I though we left terminal computing (either smart or dumb) back in the '80's. Screw that crap, I'll keep my data and aps on my own computer, thank you.

      No, ssh/VNC/X/MSTSC/"To The Cloud" (cue Hipster Victory music) is still alive and well.

    • Besides, I though we left terminal computing (either smart or dumb) back in the '80's. Screw that crap, I'll keep my data and aps on my own computer, thank you.

      Programs and data may belong to your employer and not to you.

      The terminal was a desk-bound heavyweight. The smartphone or tablet is fragile, feather-light by comparison, easily mislaid and a magnet for thieves.

      "John, the Penquin Club called to say you left your laptop behind at the bar."

      Office 365 can be bundled with a subscription or lease for full - local - install of the MS Office suite beginning at $16/mo, as I recall.

    • Re: (Score:2)

      by geekoid ( 135745 )

      You can do both, you know? You can use the cloud, and back up your data.

      Terminal computing and the cloud is the obvious move, it's just in the 90's there where still technical hurdles. For the most part those hurdles have been solved.

      The terminal computing in the 80s was just a step away from using mainframes for most peoples work.

      Different things for Different reasons.

      The cloud is a great move for most people. Remember most people don't back up anyways Now their data is on a system they can access from new

    • Re: (Score:2)

      by CAIMLAS ( 41445 )

      So-called "cloud" computing has been around for many years in one form or another, but mainly as 'hosted services'. The only difference now is that it's really, really popular (beyond what $9.95/year can do) because it looks like another computer out there on the internet, not some minimalist service. It also costs more, as you'd expect. The only thing that hasn't changed

      99.999% uptime was acquirable years ago. How do you think they did that, exactly? Magic? No - the same schemes employed now with cheaper s

    • Re: (Score:2)

      by mcrbids ( 148650 )

      Especially this early in the life cycle of this "cloud" crap. Any expectation of not loosing your data if you don't keep a backup yourself is entirely your own fault.

      Early? Come again? I've been providing "cloud computing" services for about a decade. Is 10 years early?

      Of course, the names have changed. There was "Service Oriented Architecture" and before that it was "Software as a Service" and then before that "managed application hosting" and various other names for it over time. Name doesn't matter, it's

  • Compare that with the uptime that typically have in any windows installation running the old office, for which you pay the full price, at least, if you access the new online version from a non-ms browser/operating system.

  • The cloud is horribly unreliable. (Score:5, Insightful)

    by Shoe Puppet ( 1557239 ) on Thursday June 30, 2011 @03:35PM (#36626532)

    The cloud is horribly unreliable. You should continue using Windows and Office instead.

    -- Microsoft.

  • What summary leaves out ... (Score:5, Informative)

    by MacTO ( 1161105 ) on Thursday June 30, 2011 @03:37PM (#36626544)

    ... is that more than 18 days of downtime results in a complete refund, 4 to 18 days of downtime results in a 50% refund, and 8 hours to 4 days of downtime results in a 25% refund. (Calculations are assuming 1 year of service, though I don't know how Microsoft does it.)

    This is not what I would call excellent, but it is several orders of magnitude better than the summary suggested.

    • Uhm isn't 17.999... days still about 18 days? The summary said 18 days of downtime, still pay 50%. That's a pretty accurate summary when using my kind of rounding.

      • Re: (Score:3)

        by MacTO ( 1161105 )

        The point is that the summary is horribly misleading because it is very selective about the information that it provides.

        In that respect it is even more misleading than my post since 0.05*365.242199 is closer to 18.2621099 days. You may also wish to note that there are similar rounding errors (though I'd prefer to call them conversational conveniences) in all of the figures that I presented.

        Thus my credibility is entirely destroyed and my original post should probably earn a score of "0, flame-bait" becaus

      • Just because it's accurate doesn't mean it tells the whole story. It's called "spin". Submitter clearly wants Office 365's SLA to look bad, so he focuses on the absolute least impressive number. He's written an opinion piece rather than reporting, thus making himself a pundit instead of a reporter. It's the difference between Brian Williams and Glen Beck.

        • This is a news site. Which of the following is new:

          • Microsoft won't give you a discount if it has 100% service.
          • Microsoft might give you a little discount if you have some outage.
          • If you have 18 days outage in one year, you'll still have to pay 50%
          • If they deliver almost no service you won't have to pay for it

          And come on. I'm not submitting the story because I feel it's my journalistic duty. I'm submitting it becuase I think it's interesting what Microsoft thinks it can get away with. Being a pundit

          • The fact is you're a liar, and you are going out of your way for some reason to attack one company, while leaving the other in a much better position than it deserves. You hypocritically ignored the fact that you made up Google's 100% uptime. Google doesn't even claim the figure in the article you linked and they offer similar promises.

            It's also worth noting--to you especially--that offering a discount due to downtime is not the same as guaranteeing downtime. Financially compensating for that downtime is wh

  • Slashdot... (Score:3, Insightful)

    by LBArrettAnderson ( 655246 ) on Thursday June 30, 2011 @03:43PM (#36626610)

    Come on, guys. It's just a SLA. You get a full refund if it's more than 5% downtime (18.25 days). You get half off for 99% to 95% uptime , and 25% off for 99.9% to 99%. Do you really think they're expecting to give these refunds? No. But it's there in a contract just in case. I doubt many people will even get the 25% refund. 99.9% isn't by any means terrible.

    Write an article when it actually goes down. The mindless /. MS bashing needs to stop.

    • Really? (Score:5, Interesting)

      by publiclurker ( 952615 ) on Thursday June 30, 2011 @03:55PM (#36626782)
      We use Microsoft's current online offering, and we've had both a 25% and a 50% refund in the last year and a half. the refund doesn't even begin to make up for the sales losses and confusion when our dealers can't get their orders through to us.

      • Re: (Score:2)

        by spruce ( 454842 )

        We use Microsoft's current online offering, and we've had both a 25% and a 50% refund in the last year and a half. the refund doesn't even begin to make up for the sales losses and confusion when our dealers can't get their orders through to us.

        It's always a good idea to use the cloud to host crucial business systems with no fallback plans in case there are problems. Heck, the cloud is good for anything! Personally, my "waste processing" system is routed through the cloud. Sure that two week outage last month was a bit uncomfortable, but I had considered the ramifications of this design before implementing it, as I'm sure you guys did.

      • Re: (Score:2)

        by kenh ( 9056 )

        And why did you go to the cloud if you rneeds (seemingly) can't tolerate any sort of downtime?

        I can only assume that rather than continue to put your business at risk you've since hired IT folks and created a complete, redundant datacenter that not only has redundant hardware, network access from diferent providers, and a reliable form of emergency power that can run the datacenter indefinitely so that your application can achieve the four (or more) nines your business obvioulsy demands?

        If you have not, tha

    • Still kind of sad show of confidence. Competitor offers 99.9% up-time full guarantee counting maintenance etc... Microsoft counters with a 96% partial refund guarantee. If Microsoft made the first move it would be fairly respectable, but to bring out a counter move that is less then it's target is a tad silly. Something tells me microsoft's marketing department is not getting the idea of how to one up the competition. If you come in first whatever offer you make is considered good, coming in second and offe

    • Do you really think they're expecting to give these refunds?

      Do you really think Sony expected the PlayStation Network would be down an entire month?

    • Re: (Score:2)

      by hawguy ( 1600213 )

      Come on, guys. It's just a SLA. You get a full refund if it's more than 5% downtime (18.25 days). You get half off for 99% to 95% uptime , and 25% off for 99.9% to 99%. Do you really think they're expecting to give these refunds? No. But it's there in a contract just in case. I doubt many people will even get the 25% refund. 99.9% isn't by any means terrible.

      Write an article when it actually goes down. The mindless /. MS bashing needs to stop.

      Apparently they *do* expect to have to pay out on their SLA guarantee, otherwise they would have made it a 100% refund. Or even a 150% guarantee "We're so confident in our service that we will pay *you* half your monthly fee if we're down more than 5% of the time"

      99.9% is reasonable for a cloud hosted app, but if they miss that target then they only pay me 25% of my monthly fee which doesn't really offset the cost to me if my entire workforce is idled for an hour because they can't reach their email, docum

      • Re: (Score:3)

        by kenh ( 9056 )

        Three nines is almost 9 hours of outage, but they can come in little bursts, not all at once, and still count as a total of 9 hours of outage.

        If you need three nines ( and that includes not only work hours but evening and weekend hours when your workforce probably wasn't using Live365 anyways), maybe you shouldn't be on a cloud service.

        I once worked for an employer that turned off email from 11:00 -> 1:00 every work day on purpose. It was to increase actual productivity and reduce mindless CC:'ing and en

        • Re: (Score:2)

          by hawguy ( 1600213 )

          If you need three nines ( and that includes not only work hours but evening and weekend hours when your workforce probably wasn't using Live365 anyways), maybe you shouldn't be on a cloud service.

          Somehow Google maintains nearly 4 9's of service uptime on Google apps. My internet connection (dual-homed to independent, BGP so I can use either ISP) gave me 99.95% uptime last year (including maintenance windows).

          I don't know what kind of workforce you have, but in my company I can count on people working from around 4am (from our East coast office) and ending around 11pm when people sign off for the night - this includes weekends (though obviously there are a lot fewer people working on a Saturday after

  • Frame it in the worse light possible (Score:5, Informative)

    by kervin ( 64171 ) on Thursday June 30, 2011 @03:51PM (#36626736)

    I actually thought the assurances were descent. Try looking at the SLA for your other cloud products to compare. Plus I've had Microsoft hosted Exchange for almost 2 years now and can't remember a single outage.

    But what's sad is that the title of this 'article' and summary tries so obviously and desperately to frame the SLA in the worse possible light.

    How about reporting something newsworthy, like the fact that Microsoft released Windows Phone 7.5 Beta 2 ( Mango release ) [engadget.com] to the entire development community yesterday.

  • Dear Valued Customer;

    We apologize that the recent outage has caused you to lose the multi-million dollar contract. Here is a check for $2000 to compensate for the down time.

    Sincerely,
    Microsoft

    • But would it really?

      That is to say, is your scenario that downtime of the cloud would result in the loss of a multi-million dollar contract in any way shape or form realistic?

      I am no fan of "the cloud" in this context. But is there some aspect of Office 365 (or is this now Office 347?) that would prevent people from making offline copies of their work? Wasn't the idea of the ability of making offline copies via Office 365 one of Micrsoft's earlier advantages over Google.

      The cloud may make collaboration ea

      • Re: (Score:2)

        by hawguy ( 1600213 )

        But would it really?

        That is to say, is your scenario that downtime of the cloud would result in the loss of a multi-million dollar contract in any way shape or form realistic?

        Here's a more realistic scenario: 32 hour outage spanning 2 full business days. Company is paying $24/mo for 100 users = $2400 refund. (or, 18 hour intermittent outage, spanning 2 business days)

        Company sent employees home with pay because without access to email, calendars, or documents they can't get any work done.

        Each employee costs $50/hour average so 100 * $50 * 16 hours = $80,000.

        So Microsoft paid out $2400 to "compensate" a company with a direct loss of $80K, plus indirect losses caused by the busine

        • Re: (Score:2)

          by kenh ( 9056 )

          And if it happens on the weekend, what is the "cost" of the 18 hour outage? Very nearly nothing in most cases.

          Your imagined/calculated $80K "cost" is cut in half if the outage spills over into the weekend OR spillsover from the weekend into Monday...

          Let's not forget the month over month savings from not having an IT infrastructure beyond a couple switches and an internet connection - that has to figure into your savings when weighed against the cost of an outage.

          Your imaginary "company" has 100 employees wh

          • Re: (Score:2)

            by hawguy ( 1600213 )

            And if it happens on the weekend, what is the "cost" of the 18 hour outage? Very nearly nothing in most cases.

            Your imagined/calculated $80K "cost" is cut in half if the outage spills over into the weekend OR spillsover from the weekend into Monday...

            Let's not forget the month over month savings from not having an IT infrastructure beyond a couple switches and an internet connection - that has to figure into your savings when weighed against the cost of an outage.

            Your imaginary "company" has 100 employees who suddently become incapable of producing anything of value when their Live365 access is cut off? No cached documents, no local installs of Office? There's nothing they can do? Sounds like a company that shouldn't have been on a cloud-based solution, IMHO.

            So you're telling me I get all of the cost savings from "not having an IT infrastructure beyond a couple switches and an internet connection", yet my company is stupid because I have no local installs of Office and can't get any work done without Live365?

      • Re: (Score:2)

        by kenh ( 9056 )

        Maybe, just maybe, after the first few hours of the outage it would occur to an employee to go to BestBuy and pick up a retail copy of Office and install it on their PCs? Or maybe download the tial software from MS and run that for the duration of the outage?

        Nope, I guess when a multi-million dollar contract is on the line, there's no way the employees can "think outside the box"...

      • That is why I titled the post "Timing". Just as you are about to pull together the two hour presentation the cloud goes down. You now have ten hours to pull it together. Even with off line backups there are several issues.

        1. Who has the most recent file?
        2. How many updates have been lost and need to be re-done.
        3. How do we get this presentation package together without email?

        If everyone has to keep backup copies of every part of a major presentation just in case the cloud goes down then why use the cloud? I

  • How is this a bad thing? (Score:3)

    by brainzach ( 2032950 ) on Thursday June 30, 2011 @04:12PM (#36627012)

    This forces Microsoft to put their ass on the line and deliver.

    If Microsoft risks losing half its revenue, they are going to spend the resources to prevent it.

  • Way to spin it submitter (Score:1)

    by Anonymous Coward

    From TFA

    Under the service level agreement, [Office 365] customers receive 25 per cent off their monthly payment if uptime falls below 99.9 per cent to 99 per cent, half of the sum back if it falls below 99 per cent and a complete refund for anything under 95 per cent.

    Compare this with the google apps SLA [google.com] and you'll see MSFT's is actually better.

    Monthly uptime = Credit given
    99.9% to 99.0% = 3 Day credit (MSFT gives 7.5 days)
    99.0% to 95.0% = 7 Day credit (MSFT gives 15 days)
    Under 95.0% = 15 Day credit (MSFT gives 30 days)

    It's worth noting that these are both just SLAs. There was no mention of any downtime actually happening for either service

  • Opportunity costs are the big expense (Score:3)

    by sjbe ( 173966 ) on Thursday June 30, 2011 @04:59PM (#36627608)

    The real expense isn't actually the cost of the service. The real expense is the LOST PRODUCTIVITY. That does not get compensated in form by any vendor. Frankly they could offer it for free for a year and not cover the cost of the lost productivity for a single day for a heavy office application user. 99.9% reliability means 8.76 hours of downtime per year. Someone making $20/hour would cost $175. Add in the fact that they presumably are there because their services are more valuable than their salary (otherwise why hire them?) and you can add on even more cost. Our at breakeven our company brings in revenue of about $100,000 per employee per year which for 240 working days works about to about $416/day. A seat of LibreOffice or even Microsoft Office is cheap compared with lost productivity.

    Furthermore no matter how reliable a "cloud" services vendor might be, they can never be more reliable than the internet and power connections of the customer. Getting an uptime guarantee from the ISP is not cheap and you also have to have backup power to ensure computers function when the lights go out. I've had outages where I live of several hours at least 3 times in the past 12 months.

    Cloud computing has its advantages but the economic advantages are still pretty unclear for most of us.

  • So, to take the original poster's comment that 36 hour outage in a month yields a 50% refund on service fees, that seems OK, I mean, it's 95% uptime (365-18)/365 * 100 = 95%. Would I like it if my service went down for 18 hours straight? No, of course not - but what is the suggested compensation for a 36-hour outage? 100% refund for the month despite giving the user 28 1/2 days of uninterrupted service?

    What does Google offer for a 36 hour (1 1/2 day) outage? Amazon? I suspect this is actually a generous com

  • Once someone has a day and a half of downtime they aren't going to care about getting half of their money back. Instead they will want a full refund and damages. I know in some situations, even for smaller businesses, that downtime could easily cost thousands of dollars. Unfortunately for Microsoft, there may not be a lot of people using their software in the future even if they make it free.

Slashdot Top Deals

"Protozoa are small, and bacteria are small, but viruses are smaller than the both put together."

Close