Microsoft Cloud Service is down again, this is the second time this month
On Tuesday afternoon, US time, Outlook, Hotmail, OneDrive, Skype and Xbox Live all had network failures, and global users could not sign in.
The users of Outlook mailbox system in the US coast and US coast cities are also tragically affected by the users of OneDrive in the US, Skype in Western Europe and Brazil, Xbox UK, the US and Western Europe. It was not easy for Azure users that day, and a large number of engineers could not access the system.
The Xbox, Skype and Outlook teams are the first to publish information and notify users that they have been actively repaired.
At 4:00 p.m. US time, Outlook, Hotmail, OneDrive, Skype and Xbox Live all returned to normal operation, and OneDrive is working hard to repair.
This is not the first time that Microsoft has encountered this problem recently. The last time was on March 7.
Dean Greene, senior vice president of Google, discussed the reliability of cloud services in depth. She cited the "attack" Microsoft Azure report from CloudHarmony, a Gartner consulting firm. According to the report, the downtime of Google Cloud in 2016 was 47 minutes in total. As a comparison of the downtime of Microsoft Azure service of 270 minutes, the downtime of Amazon AWS was 108 minutes.
Microsoft is refuting that because Microsoft operates larger cloud services, the downtime of the data center must be higher than Google, so downtime is not the correct way to measure the reliability of cloud services.
Microsoft said that it has 34 Azure areas worldwide, more than any other competitor (Amazon has 16 cloud service areas, while Google has only 6). They believe that the key to measuring the reliability of cloud services should be the average running time of different services in each region, rather than the total downtime. The reliability of Azure service is consistent with the measurement information of other cloud service providers. In fact, the average global operation time in the past 12 months has reached 99.9979%. For customers, only runtime is a key factor in measuring reliability.
It is undeniable that errors in the cloud system are inevitable. At the beginning of March, a cloud data center on the east coast of Amazon also failed, resulting in thousands of sites and applications being unusable, and a large number of users and companies being so confused in the world.
Microsoft's reason may be successful, but the downtime of twice a month is not optimistic, which does no harm to the reputation of Microsoft's cloud services.
