Macworld Forums

Macworld Forums: A short history of cloud computing outages - Macworld Forums

Jump to content

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

A short history of cloud computing outages

#1 User is offline   Macworld 

  • Story Poster
  • Group: MW Bot
  • Posts: 31,680
  • Joined: 30-November 07

Posted 12 October 2009 - 01:00 PM

Post your comments for A short history of cloud computing outages here
0

#2 User is offline   dreyfus 

  • Member
  • Group: Macworld Insiders
  • Posts: 975
  • Joined: 05-January 06

Posted 12 October 2009 - 01:22 PM

Well, just to put this a bit into perspective, the latest Gartner study (I have seen) about in-house server outages in corporations was putting the bar close to 90 hours/year... So, each and every example provided here is better than what dedicated IT departments are achieving?!

Without having any hard data or evidence for this ambitious thesis: I suspect a significant amount of in-house backups, at least for e.g. databases and groupware servers, would prove worthless if ever needed.

I see data security and confidentiality as two huge question marks when discussing "cloud computing", especially when these services are outsourced and the company outsourcing the work is obviously unable to make the required specifications and set and enforce suitable standards. Availability was not a big concern until now and, with the exception of the recent and expected MS goof-up, data loss was not frequent either.
0

#3 User is offline   rab777hp 

  • Veteran
  • PipPipPip
  • Group: Banned
  • Posts: 1,840
  • Joined: 11-June 08

Posted 12 October 2009 - 01:31 PM

Very insufficient article.
0

#4 User is offline   cv 

  • Member
  • PipPip
  • Group: Members
  • Posts: 384
  • Joined: 08-December 06

Posted 12 October 2009 - 02:12 PM

View Postrab777hp, on 12 October 2009 - 01:31 PM, said:

Very insufficient article.


It's not supposed to be a comprehensive, exhaustive listing of every cloud computing meltdown. After all, the title is "A short history of cloud computing outages."

That said, the author probably should have mentioned MobileMe's embarrassingly awful launch in July 2008.
0

#5 User is offline   downunder 

  • Member
  • PipPip
  • Group: Members
  • Posts: 211
  • Joined: 30-September 04

Posted 12 October 2009 - 06:11 PM

View Postrab777hp, on 12 October 2009 - 01:31 PM, said:

Very insufficient article.

Very insufficient response.
0

#6 User is offline   leskern 

  • Member
  • Group: Macworld Insiders
  • Posts: 121
  • Joined: 10-March 09

Posted 12 October 2009 - 06:12 PM

Cloud computing is just trying to remind people that the very concept is deeply flawed. First it depends on an internet connection, and in THIS country it's pretty much a crap-shoot that you'll be up for 24 straight hours. Second, it relies on ever-increasingly complex systems... almost designed to fail. Third, and most important, it relies on PEOPLE. Three strikes and we're all out.
0

#7 User is offline   dmjossel 

  • Member
  • PipPip
  • Group: Members
  • Posts: 48
  • Joined: 17-May 06

Posted 12 October 2009 - 08:19 PM

What we need is a Discounted Traffic Flow model. The value of data is inversely proportional to your distance from it, but proportional to the number of locations that can connect to it.
0

#8 User is offline   Splashman 

  • Member
  • PipPip
  • Group: Members
  • Posts: 77
  • Joined: 27-September 05

Posted 13 October 2009 - 12:01 AM

Notice that all the listed examples are "service outages" -- meaning, the service was unavailable for some period of time. None of the examples involved actual data loss, such as the Sidekick incident.

This truly is unprecedented, so anyone wanting to pin the problem on cloud computing in general is barking up the wrong tree. All this proves is that Microsoft is managed by morons.
0

#9 User is offline   henryhbk 

  • Member
  • PipPip
  • Group: Members
  • Posts: 175
  • Joined: 08-October 07

Posted 13 October 2009 - 05:04 AM

View Postleskern, on 12 October 2009 - 06:12 PM, said:

Cloud computing is just trying to remind people that the very concept is deeply flawed. First it depends on an internet connection, and in THIS country it's pretty much a crap-shoot that you'll be up for 24 straight hours. Second, it relies on ever-increasingly complex systems... almost designed to fail. Third, and most important, it relies on PEOPLE. Three strikes and we're all out.


Really? by "THIS" I assume you mean the US? Depends where you live. I have FIOS, and can't remember the last hiccup... I get reliable 20mbps service. Also http connections are fairly tolerant of brief blips in connection, since they are stateless, and since I develop cloud applications in MENA (Middle-East, North-Africa), let me tell you, you have no idea how fault tolerant a web connection is until you've played in that sandbox...
0

#10 User is offline   flowney 

  • Member
  • PipPip
  • Group: Members
  • Posts: 625
  • Joined: 14-March 04

Posted 13 October 2009 - 08:35 AM

This piece just begs for colorful charts and graphs.
Dr. Frank Lowney Georgia College & State University
Senior Director for External Projects
and Assistant to the Director, Digital Innovation Group @ Georgia College
0

#11 User is offline   Photog_7 

  • Member
  • PipPip
  • Group: Members
  • Posts: 16
  • Joined: 16-March 09

Posted 13 October 2009 - 08:40 AM

View Postdreyfus, on 12 October 2009 - 01:22 PM, said:

. . . I see data security and confidentiality as two huge question marks when discussing "cloud computing", especially when these services are outsourced and the company outsourcing the work is obviously unable to make the required specifications and set and enforce suitable standards. Availability was not a big concern until now and, with the exception of the recent and expected MS goof-up, data loss was not frequent either.


You are right about that! In this economy, you might open the newspaper (if there are still any left) and read that the company storing your data is no longer answering the phone and appears to have disappeared. Several medical practices in my vicinity had many terabytes of medical data stored at a facility in another state only to find out that the data storage company was going out of business. It took them more than 3 months to get their data, and they were lucky to get it at all. What a mess! Now, the doctors where I work are unwilling to use ANY off-site hosting or storage except for an off-site backup. They have purchased triple-redundant servers and now refuse to let anything except the Internet connection be outsourced. I've been trying to come up with industries where temporary outages and or losses would not lead to a serious crisis and concluded that cloud computing would best be reserved for recreational uses like music, movies, and games. At the clinic where I work, not having access to a medical record or a medical image is not just an inconvenience. The current administration is pushing the entire medical industry to rush into electronic medical records, and there are many new start-up companies out there pushing Internet-hosted services and storage to doctors and clinics, often from facilities located on the opposite side of the country which no one will ever see. I'm pretty sure we will be hearing about all kinds of serious problems if the medical community embraces cloud computing.
0

Share this topic:


Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users