ie8 fix
madison

What Microsoft's online outage says about its cloud strategy

By | May 15, 2011, 2:37pm PDT

Summary: Last week, Microsoft had four separate issues that knocked its cloud-based e-mail services offline for up to 9 hours. Those incidents highlight the most important thing to keep in mind when buying any cloud service.

When I wrote last week about the dangers of Google’s cloud-only strategy, a handful of commenters criticized me for focusing on the failure of Google’s Blogger service and not mentioning Microsoft’s own cloud-related problems. One questioned why I should “waste time ripping on Google, when Microsoft (your apparent reason for blogging) gets a pass from you from their arguably more annoying outage.”

Talk about missing the point.

Yes, Microsoft’s outage last week was annoying and potentially costly to paying customers. If you’re a current or prospective customer of Microsoft’s Business Productivity Online Services (BPOS), you’ll want to look carefully at how the company handled last week’s outages and what their response says about the long-term reliability of BPOS.

The fact that an outage happened shouldn’t be a surprise. In fact, it actually supports my argument against putting everything you own in the cloud.

Any network-based service can have an outage. Do you know any online service that offers a 100% uptime guarantee? Me neither.

But hard disks and file servers can fail, too, which is why I recommend a combination of local and cloud-based storage. As I wrote last week:

If your data matters, you need a hybrid strategy, with local storage and local content creation and editing tools. If your local storage fails, you can grab what you need from the cloud. If your cloud service fails, you’ve still got it locally. But if you rely just on the cloud, you’re vulnerable to exactly this sort of failure.

The troubles with BPOS were well covered here at ZDNet. Shortly after my post went live, our editor in chief, Larry Dignan, posted Microsoft BPOS uptime sparking customer angst, pleas for help. Mary Jo Foley followed that up with Microsoft’s BPOS cloud customers hit by multi-day email outage a little later in the day. In a blog post at the end of the day, Microsoft Corporate VP Dave Thompson acknowledged that the Online Services division he runs had four separate service issues that affected its hosted Exchange service:

  • An outage on Tuesday resulted in e-mail backlogs and delays that lasted up to 9 hours for some customers.
  • Two more episodes on Friday lasted 45 minutes and 3 hours, respectively.
  • A separate DNS issue early Friday morning affected Outlook Web Access, with a lesser impact on Outlook and some Exchange ActiveSync devices. That episode lasted nearly four hours.

So what makes this outage different from, say, the more-than-30-hour outage that wiped out 40,000 Gmail accounts back in February?

Update: In the Talkback section, some commenters suggest that this is an unfair comparison. They mistakenly believe that the Google outage applied only to the free Gmail service. That is not true. The official Google incident report (PDF) makes clear that this issue affected paying customers in the Google Apps for Business program, who lost all access to their data for a period of at least 32 hours.

Google has no offline client. All of those Gmail customers eventually got their data restored, but for a day and a half, they were out of luck. They had no calendar details and no contacts except those that had been synced to a smartphone, and no e-mail messages except those they had manually configured for use with their own e-mail client using IMAP.

Microsoft, by contrast, has Outlook, which is designed to keep a fully synchronized offline copy of everything in your Exchange account. Exchange and Outlook are designed to work together, and in fact a copy of Outlook is part of the system requirements for both BPOS and Office 365. In its FAQ for Office 365, Microsoft notes that “Outlook 2007 and Outlook 2010 have been designed to work with cloud services and support the cloud services’ architecture.”

BPOS customers last week had to use alternate accounts to send e-mail, but they didn’t lose access to calendars, contacts, and e-mail archives.

That is exactly what I’m talking about when I refer to a hybrid service. If I were a BPOS customer, I’m sure I would have been extremely unhappy last week. But I wouldn’t have been unproductive.

Ironically, the successor to BPOS, Office 365, is now in beta testing. It stayed online all week while BPOS was giving it customers an online roller-coaster ride. In my first look at Office 365, the first bullet point on my list of things that make it worth a long look is its “great online/offline support.” That’s exactly the point I make here.

Microsoft does offer cloud-only online services of its online services. The BPOS deskless worker options and the equivalent Kiosk Worker plans for Office 365 both deliver web-only e-mail packages at a reduced price. But those packages are intended for workers who share a terminal and typically spend only 5-10% of their time at a PC. The Enterprise plan includes “full client connectivity.”

When it comes to the cloud, Microsoft says they’re “all in.” Thank goodness that’s just a slogan.

Related stories:

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

Topics

Ed Bott is an award-winning technology writer with more than two decades' experience writing for mainstream media outlets and online publications.

Disclosure

Ed Bott

Ed Bott is a freelance technical journalist and book author. All work that Ed does is on a contractual basis.

Since 1994, Ed has written more than 25 books about Microsoft Windows and Office. Along with various co-authors, Ed is completely responsible for the content of the books he writes. As a key part of his contractual relationship with publishers, he gives them permission to print and distribute the content he writes and to pay him a royalty based on the actual sales of those books. Ed's books written prior to fall 2011 have been distributed by Que Publishing (a division of Pearson Education) and by Microsoft Press. As of November 2011, Ed is a partner in the independent publishing company Fair Trade Digital Exchange, which exclusively publishes his books.

On occasion, Ed accepts consulting assignments. In recent years, he has worked as an expert witness in cases where his experience and knowledge of Microsoft and Microsoft Windows have been useful. In each such case, his compensation is on an hourly basis, and he is hired as a witness, not an advocate.

Ed does not own stock or have any other financial interest in Microsoft or any other software company. He owns 500 shares of stock in EMC Corporation, which was purchased before the company's acquisition of VMware. In addition, he owns 350 shares of stock in Intel Corporation, purchased more than two years ago. All stocks are held in retirement accounts for long-term growth.

Ed does not accept gifts from companies he covers. All hardware products he writes about are purchased with his own funds or are review units covered under formal loan agreements and are returned after the review is complete.

Biography

Ed Bott

Ed Bott is an award-winning technology writer with more than two decades' experience writing for mainstream media outlets and online publications. He's served as editor of the U.S. edition of PC Computing and managing editor of PC World; both publications had monthly paid circulation in excess of 1 million during his tenure. He is the author of more than 25 books on Microsoft Windows and Office, including the recently released Windows 7 Inside Out.

Related Discussions on TechRepublic

Did you know you can take part in these discussions with your ZDNet membership?
135
Comments

Join the conversation!

Just In

RE: What Microsoft's online outage says about its cloud strategy
FAULKNE 13th Oct
Good day to confirm this comment I would appreciate T h e b e s t o f Z D N e t d e l i v e r e d your website very nice to everyone Yes, Oracle is the only one with shared-disk architecture, but that is there advantage. It means you can add or remove nodes and the database lives on. In a shared nothing architecture, if you lose a node, you lose the system. I'm sure Oracle appreciates EMC highlighting their advantage.I also desire to signal in your RSS feeds. Thank you as soon as once again and maintain up the great operate Awesome post! Thank you very much || thanks for nice content this is really benefit to me.
0 Votes
+ -
Excellent post
facebook@... 15th May 2011
I would also like to hear you explore the differences in the SLA. A google outage results in more time being added to the end of the contract. A microsoft outage results in remuneration. Another key differentiator in their approaches.
@facebook@...

SLA is important, but at least somewhat irrelevant when your service is down. We recently had an Internet outage, and there's not much the ISP could have done to compensate us for the lost produtivity during the 2-3 business days involved. The business-class SLA didn't prevent the outage or their incompetence in fixing it, let alone our time and cost to put in a redundant connection from another provider.
@DaveN_MVP

What are the penalties that your carrier provides in the event of a downtime? I tend to deploy dual carriers for companies that require a high level of uptime. YMMV. However, it sounds like you did not align your continuity needs with your vendor SLAs. A central part of managing any vendor relationship is to ensure that your true business needs are clearly understood before commencing SLA requirements.
0 Votes
+ -
@facebook@... At $2.50 per user for a problem that persisted for a month, do you feel that companies were adequately compensated?

I am surprised that more Microsoft Partners are not speaking up in this forum. Perhaps they are too busy trying to explain how an email issue could last so long and why they were powerless to do anything to fix it.
@MaranathaP

http://microsoftvolumelicensing.com/DocumentSearch.aspx?Mode=3&DocumentTypeId=37

A 100% credit is extremely compelling, as compared to additional days at the tail end of a controal.
You are welcome and thank you! ^_^ rolex airking replica
I'm still amused at the notion that M$ came to market with a product acronym "BPOS".
0 Votes
+ -
Contributr
Be consistent
Ed Bott 15th May 2011
@Non-techie Talk

If you're gonna trash-talk, then don't you need to spell it BPO$?

wink
@Ed Bott
LOL. grin
@Ed Bott: ... will be out for at least one day every four years?

Microsoft indeed has really lame naming habits; always had.
@Ed Bott Touche! I appreciate your having fun with this.
@Non-techie Talk They didn't... it was called Business Productivity Online Services when it was released.
@Non-techie Talk
@JonathonDoe -- Now that's hilarious!!!!!
@JonathonDoe LOL!
0 Votes
+ -
Good analysis
LiquidLearner 15th May 2011
It was delayed e-mails. E-mails weren't lost, deleted, etc. You still had access to all the online services. So it wasn't a total outage, but it was very frustrating. Given the fact I had just switched a client from a failing Exchange server over the weekend. Oh well, it's how online services go. That said, in the two years I've been supporting users on BPOS this is the first outage that lit up my phone. For the most part it's been incredibly reliable.

I think they easily meet the uptime SLA overall but on a monthly basis, this month they fail. Good thing people still have IM, other e-mail, and phones.
0 Votes
+ -
This went on for a month
MaranathaP 16th May 2011
@LiquidLearner The meltdown might have occurred over a 2 day period, but there were ongoing mail flow issues for a month prior to this. Every ticket we opened received the following generic response: "Concerning Service Request xxxxxx-xxxxxx, the BPOS Operations team continues to monitor email flow in the environment. The team has resolved issues with users seeing email stuck in draft or outbox. Users with mail that was previously queued in their outbox may see delays of up to 3 hours for final delivery based on the significant amount of email that is queued. Microsoft is closely monitoring the situation and expects mail queues to reduce to normal levels in the next several hours."
This was a month long problem for all of our clients in the US.
0 Votes
+ -
Missing the point indeed
ego.sum.stig@... 15th May 2011
You have. You ignored the BPOS outage (a Microsoft issue), then ranted about Google's (not a Microsoft issue) and now you're defending BPOS's outage and saying it isn't as bad as Google's.

You are apparently paid write a "Microsoft Report" blog. Who knew that Google were Microsoft?
0 Votes
+ -
Contributr
@ego.sum.stig@...

My focus is on the customer... You seem to be more interested in proving some sort of point.
0 Votes
+ -
@Ed Bott :
"one of its flagship services has failed spectacularly."
now look at the soft talk you use for BPOS issues...
" Microsoft?s outage last week was annoying" then you give a sales pitch for the MS products.
Thats why you were pulled up on trashing google..
@Ed Bott
@ego.sum.stig

I support e.s.s on this as you compared two completely different products. The Gmail downtime was for a free service, you should have considered the paid service for comparison, and yes I got the same feeling as others here that you prefer one more than the other...
0 Votes
+ -
Contributr
Sorry, that is not correct
Ed Bott 16th May 2011
@athlonica

The Google outage I described affected paying customers as well. I have updated the post to include a link to the official Google incident report, which confirms that 40,000 customers lost ALL access to e-mail, calendar, and contacts for at least 30 hours.

The BPOS outage last week involved delays in sending and receiving e-mail of up to 9 hours. There is a difference.
0 Votes
+ -
102% agree with u
aliasadadm 16th May 2011
@Ed Bott
Ur point is most valid as far as 100% cloud dependance is concerned. U r right & s far as my knowledge no company can guarantee flawless 100% uptime and that is the biggest reason not to go on cloud 9. Sometimes these commenters are unbelievable, may be paid by Google.
0 Votes
+ -
@Ed Bott There has not been, as far as I am aware, an actual tally of how many users were affected by Microsoft's ongoing outage. With how hard they have been pushing Partners to sell BPOS, I would dare guess it far greater than 40,000 people.
In addition, this problem was ongoing for a lot longer than has been reported with message flow delays for about a month.
Microsoft is Partner driven and WE are the one's who have to face the repercussions.
It has been very costly to us indeed.
0 Votes
+ -
Perhaps the ones who missed the point was YOU
Pete "athynz" Athens 15th May 2011
@ego.sum.stig@... What part of Ed's conclusion about a hybrid cloud/ local system being better than either total cloud or total local was lost on you? He was using Google's outage as an example as the MS outage had been covered by several other bloggers as indicated in his post here. But let me guess - you didn't bother to read either article.

Ed thinks you are trying to prove a point, I think you are some sort of troll who can't be bothered to read beyond the article's headline - or byline.
0 Votes
+ -
Ed has a long history...
jasonp@... 16th May 2011
@athynz
of downplaying or completely ignoring problems in Redmond while pointing out problems from everywhere else. He also has this odd habit of comparing apples and oranges, like his attempt to compare BPOS and Google Mail in this article. I don't know anyone who pays to use Google Mail, do you? Finally, he has a long history of using "truthiness" to make points. Like his assertion that Google Mail doesn't have an offline client. While technically true, the statement doesn't paint a truthful picture. I and many people I know use Outlook as a Google Mail client. I know other people who use different offline clients to access Google Mail. Ed throws that little tidbit out there hoping people will read into it that Google Mail doesn't support an offline client. That would be like me claiming Microsoft Exchange does not have an offline client, which is technically true but intellectually dishonest.
0 Votes
+ -
Contributr
You're kidding, right?
Ed Bott 16th May 2011
@jasonp

"I don't know anyone who pays to use Google Mail, do you?"

Yes. In fact, the outage I reference here was for Google Apps for Business:

http://static.googleusercontent.com/external_content/untrusted_dlcp/www.google.com/en/us/appsstatus/ir/nfed4uv2f8xby99.pdf
0 Votes
+ -
Amusement abounds
ego.sum.stig@... 16th May 2011
I thought the point of posting was to make a point. Actually there were a few I tried to make.

Ed seems to be trying to make a point that cloud only is a bad idea. But that thesis is voided for the most part due to the waste of time and space trashing one cloud provider and excusing another. His blog is supposed to be a "Microsoft Report." Is it too much to ask that he give a Microsoft report?

I'd have rather read what he thought about the ins and outs of Microsoft's cloud thingies. After all, he does appear to know Microsoft stuff fairly well.

However, and it's not just him, bloggers seem to write whatever in whatever manner that will generate the most page hits and forget actually being informative or even slightly useful.
  • Flagged
0 Votes
+ -
Message has been deleted.
jasonp@... Updated - 17th May 2011
  • Flagged
@ego.sum.stig@... As athynz said, you seem to have missed the point.

I've been saying for years, that a pure cloud approach doesn't make sense, especially for businesses. A hybrid approach is the only sensible option, which is exactly what Ed was talking about here.

Yes, it put MS is a better light than Google, but that is because of their background. Google are a web platform first and foremost, they understand web and they want everybody to work on their services in the web...

Microsoft come from an in-house background, where everything was local, then on servers inside the firewall, now they are trying to keep pace and move into web platforms.

Ironically, it is their background in local apps which gives them such an advantage in the cloud at the current time.

It doesn't have to be the cloud provider that has problems, if the ISP has problem, or like us, your firewall dies, you are stranded, if you only work in the cloud.

We could continue working, until our internet connection was restored. The only things that didn't work were e-mail and remote support, we had to go back to the old ways of remote controlling the user, as opposed to remote controlling their computer.
0 Votes
+ -
The old ways of remote controlling the user
Will Pharaoh 16th May 2011
@wright_is
Funny, but very accurate!
  • Flagged
0 Votes
+ -
@ego.sum.stig@...
but say little when it happens to someone other then MS.

So what's your angle in all of this? Google gets a pass but MS shouldn't?
0 Votes
+ -
Any unplanned outage is BAD ...
mwagner@... 16th May 2011
... but loss of data is ALWAYS worse than loss of service and the longer the service outage, the worse it gets. Ed's point is that Microsoft's service is focused on the needs of the enterprise and the impact of their outage reflects that. The impact of Googles outage reflects that Google just doesn't understand the needs of the enterprise customer. it also indicates that Google's offering is not as robust as Microsoft's offering.
@ego.sum.stig@...
Unbelievable, may be u r one of those people who say "Google pays me 60$ an hour"
0 Votes
+ -
FYI
ego.sum.stig@... 16th May 2011
I don't work for Google. I torture mainframes for a living. Ergo, cloud stuff is of some interest to me. It would be nice to read stuff that wasn't all trash and bash.
  • Flagged
@ego.sum.stig@... I'd recommend you'd actually read what he's saying. You are jumping the gun and have not understood Jack.

Short: If you are 100% cloud-based (no matter who with) and your Cloud goes down, you are SOL. There is NOTHING you can do. Productivity will be at exactly 0. You have ONE option, and that is to twiddle thumbs until your Cloud provider is back up. You can not affect the time that will take.

If you are 0% Cloud, you are in control of what happens when things go pear-shaped. If you have a back-up strategy and, in case of total wipe-out, a plan for getting new hardware hardware, you decide how long the outage is going to be.

The advantages of the Cloud are many though, so Bott is advocating you go with a hybrid strategy, basically keeping your in-house and your Cloud in sync. This means that when the Cloud goes down your productivity will suffer significantly less than it otherwise would.

Botts piece had nothing to do with Google or Microsoft, it was about how you, as a company, can create a strategy for how to minimize down-time.

To me, honestly, even considering putting all of my stuff into the cloud, and ONLY in the cloud is absurd and insane. Any IT professional that suggested it would be terminated on the spot. The Cloud is great, and it increases my productivity, but I'm not going to leave critical data in the cloud only. The cloud will go down. When it does, I have all my data locally. Backed up and secure.
0 Votes
+ -
Reading
ego.sum.stig@... 19th May 2011
A lost art. But not entirely, I do read, I did read.

It is amusing that Ed decided to froth at the mouth about being called on:
a) The title of his blog and why he feels the need to write flamebait blogs on anything but Microsoft.
b) Why he is not as harsh with Microsoft's BPOS outage which has been more like a painful rolling brownout than a single event, yet feels the need to declare catastrophe for Google for what appears to be a single event.

Other than that, it's kind of amusing why zdnet bloggers are more and more tending to froth and sensation rather than thought and depth.
0 Votes
+ -
Does Office 365 has a offline client ?
jinishans Updated - 16th May 2011
I've registered for the trial, but not sure about the offline client option for Office 365.

Update: Thanks for all of you.
@jinishans The Offline client is Office. Preferably Office Pro since it comes with SharePoint Workspace. It's N-1 so if you run 2010 or 2007 Office it is support.
@jessiethe3rd

You can built your own cloud based apps using simple tools like ThinServer

http://www.aikotech.com/thinserver.htm
  • Flagged
@Jen8 Or I can just use App-V in the Desktop Optimization Pack... Or I can use RDP... Or I can have it hosted in a partner's datacenter with Software Provider Licensing Agreement (SPLA)
@jinishans The basic model is cloud-only, the premium model includes a copy of MS Office for off-line/serious work.
0 Votes
+ -
Actually based on the record they are worst than the other providers ... but I will give them the benefit of the doubt and just say that they are just about the same level of crappy as everyone else.
@wackoae based on what record? do you have a link that compares uptime of major cloud-based productivity suite providers?
@wackoae
>>Actually based on the record they are worst than the other providers ...
Could you please back it up?
Or do you really think that is acceptable?
0 Votes
+ -
Contributr
Source? Link?
Ed Bott 15th May 2011
@wackoae

You realize that that is the defiition of 99.8% uptime?
0 Votes
+ -
You should know
wackoae Updated - 15th May 2011
@Ed Bott Check back on your own blogs.
@wackoae So are contesting that you can keep a mail server up past 99.9% uptime for let's say 100 users at a lower cost? Please make sure you factor in storage cost, hardware cost, people wage cost...
0 Votes
+ -
Contributr
You gonna keep playing games?
Ed Bott 16th May 2011
@wackoae

I went back and "checked my own blogs." I have no idea what your are talking about. Put up or shut up, please.
Good day to confirm this comment I would appreciate T h e b e s t o f Z D N e t d e l i v e r e d your website very nice to everyone Yes, Oracle is the only one with shared-disk architecture, but that is there advantage. It means you can add or remove nodes and the database lives on. In a shared nothing architecture, if you lose a node, you lose the system. I'm sure Oracle appreciates EMC highlighting their advantage.I also desire to signal in your RSS feeds. Thank you as soon as once again and maintain up the great operate Awesome post! Thank you very much || thanks for nice content this is really benefit to me.

Join the conversation!

Formatting +
BB Codes - Note: HTML is not supported in forums
  • [b] Bold [/b]
  • [i] Italic [/i]
  • [u] Underline [/u]
  • [s] Strikethrough [/s]
  • [q] "Quote" [/q]
  • [ol][*] 1. Ordered List [/ol]
  • [ul][*] · Unordered List [/ul]
  • [pre] Preformat [/pre]
  • [quote] "Blockquote" [/quote]
ie8 fix
ie8 fix

The best of ZDNet, delivered

ZDNet Newsletters

Get the best of ZDNet delivered straight to your inbox

Facebook Activity

White Papers, Webcasts, & Resources
ie8 fix
ie8 fix