January 31, 2013 at 16:50Most Notable Cloud Provider Issues for 2012
Most recently ended 2012. And it ended with very controversial results for Cloud Technologies. Many lost money, time and nerves in different situations caused by providers. And providers, in turn, have spoiled their reputation.
In my opinion, it is too early to talk about complete trust in Cloud technologies and the high stability of such services. In other matters, this is a topic for discussion; there can be no unequivocal opinion here.
I suggest you look at the 10 largest failures and problems of cloud providers in 2012 and draw conclusions.
In 2012, problems and failures plagued cloud providers with enviable persistence, giving the impression that a failure is not an event that goes beyond the normality, but quite the contrary, is an everyday occurrence.
This situation contributed to negative sentiment. Probably, you should not yet trust your core assets to public cloud providers. Accordingly, hybrid and private clouds have become more widespread and developed. Nevertheless, let's look at the direct participants in this drama:
10. Tumblr The
blogging platform did not function for several hours on December 3, due to the anti-blogging group 'GNAA' attack. According to some reports, more than 8,600 users were affected. Tumblr himself did not explain the causes of the problems, however, he notified users that their personal information remained safe and sound and did not fall into third parties.
9. GoDaddy
September 10 at 6 o’clock GoDaddy became unavailable. This company has a huge number of customers, both large companies and small sites. There is no information about who lost how much. The company claimed that the problems were caused by a series of internal network errors, and not by an anonymous group of hackers as intended. A month later, GoDaddy announced that they are turning off the provision of cloud hosting and switch to providing other services.
8. Salesforce.com
Some of the Salesforce.com services did not work on July 10 for reasons related to the failure of new equipment in West Coast data centers. For some customers, services were unavailable for up to 2 days!
Earlier, on June 28, an error in the software caused problems with data warehouses, which also served as a negative factor.
7. Dropbox
In October, the 26th, problems continued for several hours. The error message on the site looked like this: “Error: Something went wrong. “Don’t worry, your files are still safe and the Dropboxers have been notified.”
Interestingly, on the same day, Google App Engine as well as Tumblr had problems. All this gave rise to the opinion that the problems with Internet access were global, however, no one knows the real reasons and they have not been established.
6. Google Talk The
popular and sought after instant messaging service did not work for almost 5 hours on July 27th.
The news feed of the service was constantly updated and kept users informed of what was happening, later Google apologized immediately after the service was restored. The apology said: “The main priority for Google is the availability of services and we are working to constantly improve stability.”
5. Google App Engine
On October 26, Google App Engine, a platform for hosting and application development, was unavailable for almost 4 hours. As a result, approximately 50% of server requests did not reach the final goal. Later, the company reported that the data was not lost and the normal functioning of the service was restored. The following measures were also taken: “We increased the traffic capacity to avoid similar situations in the future.”
4. Microsoft Office 365
In November, 2 times, Microsoft Office 365 users for some time were unable to use email services.
“We all here understand what responsibility we bear before our users and we realize how many problems were created for the business of our customers. This is unacceptable, ”said Rajesh Jah, vice president of Microsoft Office for his blog.
Customers in the Americas could not use the service for around 8 hours, November 8th. On November 15, the situation repeated itself, this time the problems lasted for 5 hours.
3. Microsoft Windows Azure
Due to problems with hardware configuration and as a result of loss of traffic in one of the regions of Eastern Europe, Microsoft Windows Azure was unavailable for 2.5 hours. This happened on July 26th. The consequences of such an incident are difficult to predict and evaluate.
2. Microsoft Windows Azure. Again.
February 28-29, Microsoft Windows Azure suffered globally and seriously, discontinuing the service for more than a day. The reason was the problem, later called the “bug of a leap year”. The time was considered wrong for a leap year, thereby invalidating the certificates used for authentication. Some of the users were outraged and claimed that Microsoft could not adequately respond to the problem. “I noticed that there are no approximate deadlines for resolving the problem. What is the reason for this behavior? ” - noticed one of the users.
1. Amazon Web Services. Again and again.
Amazon was hit twice in 2012. In 2011, this happened 1 time. On June 14, 2012, problems with services in Virginia stopped work for more than 6 hours, affecting customer data.
On October 22, Amazon crashed in Northern Virginia, making sites hosted in the region unavailable. Even an approximate number of such is impossible to imagine. These problems have shown that even the leading provider of cloud technologies is not able to ensure 100% availability of their services.
So, problems, problems, problems. HA providers' promises are actually just promises. At least for now. So it’s still too early to talk about complete trust, security and reliability of data in the cloud.
At the same time, you can at least try to protect yourself from these kinds of problems. To do this, use your own Fault Tolerance and Disaster Recovery solutions. Also, do not skimp and use Multi-Zone Deployment in your applications. Quality services are the most important thing in our work. And you need to depend on someone as little as possible.
Let's see what will happen in 2013. Thank you and good luck to all of us.
In my opinion, it is too early to talk about complete trust in Cloud technologies and the high stability of such services. In other matters, this is a topic for discussion; there can be no unequivocal opinion here.
I suggest you look at the 10 largest failures and problems of cloud providers in 2012 and draw conclusions.
In 2012, problems and failures plagued cloud providers with enviable persistence, giving the impression that a failure is not an event that goes beyond the normality, but quite the contrary, is an everyday occurrence.
This situation contributed to negative sentiment. Probably, you should not yet trust your core assets to public cloud providers. Accordingly, hybrid and private clouds have become more widespread and developed. Nevertheless, let's look at the direct participants in this drama:
10. Tumblr The
blogging platform did not function for several hours on December 3, due to the anti-blogging group 'GNAA' attack. According to some reports, more than 8,600 users were affected. Tumblr himself did not explain the causes of the problems, however, he notified users that their personal information remained safe and sound and did not fall into third parties.
9. GoDaddy
September 10 at 6 o’clock GoDaddy became unavailable. This company has a huge number of customers, both large companies and small sites. There is no information about who lost how much. The company claimed that the problems were caused by a series of internal network errors, and not by an anonymous group of hackers as intended. A month later, GoDaddy announced that they are turning off the provision of cloud hosting and switch to providing other services.
8. Salesforce.com
Some of the Salesforce.com services did not work on July 10 for reasons related to the failure of new equipment in West Coast data centers. For some customers, services were unavailable for up to 2 days!
Earlier, on June 28, an error in the software caused problems with data warehouses, which also served as a negative factor.
7. Dropbox
In October, the 26th, problems continued for several hours. The error message on the site looked like this: “Error: Something went wrong. “Don’t worry, your files are still safe and the Dropboxers have been notified.”
Interestingly, on the same day, Google App Engine as well as Tumblr had problems. All this gave rise to the opinion that the problems with Internet access were global, however, no one knows the real reasons and they have not been established.
6. Google Talk The
popular and sought after instant messaging service did not work for almost 5 hours on July 27th.
The news feed of the service was constantly updated and kept users informed of what was happening, later Google apologized immediately after the service was restored. The apology said: “The main priority for Google is the availability of services and we are working to constantly improve stability.”
5. Google App Engine
On October 26, Google App Engine, a platform for hosting and application development, was unavailable for almost 4 hours. As a result, approximately 50% of server requests did not reach the final goal. Later, the company reported that the data was not lost and the normal functioning of the service was restored. The following measures were also taken: “We increased the traffic capacity to avoid similar situations in the future.”
4. Microsoft Office 365
In November, 2 times, Microsoft Office 365 users for some time were unable to use email services.
“We all here understand what responsibility we bear before our users and we realize how many problems were created for the business of our customers. This is unacceptable, ”said Rajesh Jah, vice president of Microsoft Office for his blog.
Customers in the Americas could not use the service for around 8 hours, November 8th. On November 15, the situation repeated itself, this time the problems lasted for 5 hours.
3. Microsoft Windows Azure
Due to problems with hardware configuration and as a result of loss of traffic in one of the regions of Eastern Europe, Microsoft Windows Azure was unavailable for 2.5 hours. This happened on July 26th. The consequences of such an incident are difficult to predict and evaluate.
2. Microsoft Windows Azure. Again.
February 28-29, Microsoft Windows Azure suffered globally and seriously, discontinuing the service for more than a day. The reason was the problem, later called the “bug of a leap year”. The time was considered wrong for a leap year, thereby invalidating the certificates used for authentication. Some of the users were outraged and claimed that Microsoft could not adequately respond to the problem. “I noticed that there are no approximate deadlines for resolving the problem. What is the reason for this behavior? ” - noticed one of the users.
1. Amazon Web Services. Again and again.
Amazon was hit twice in 2012. In 2011, this happened 1 time. On June 14, 2012, problems with services in Virginia stopped work for more than 6 hours, affecting customer data.
On October 22, Amazon crashed in Northern Virginia, making sites hosted in the region unavailable. Even an approximate number of such is impossible to imagine. These problems have shown that even the leading provider of cloud technologies is not able to ensure 100% availability of their services.
So, problems, problems, problems. HA providers' promises are actually just promises. At least for now. So it’s still too early to talk about complete trust, security and reliability of data in the cloud.
At the same time, you can at least try to protect yourself from these kinds of problems. To do this, use your own Fault Tolerance and Disaster Recovery solutions. Also, do not skimp and use Multi-Zone Deployment in your applications. Quality services are the most important thing in our work. And you need to depend on someone as little as possible.
Let's see what will happen in 2013. Thank you and good luck to all of us.