News

Microsoft Azure Recovering from Widespread Outages

Widespread "partial disruptions" affecting Microsoft's' Azure cloud computing services Tuesday night are now "mitigated," according to Microsoft.

Microsoft sent the following Twitter message late Tuesday evening:

As of Wednesday morning, the Azure status page indicated that "Azure core platform components are working properly," although Microsoft was still investigating issues affecting Azure Virtual Machine users in West Europe.

The Azure service disruptions hit Microsoft's European service regions, as well as regions in the United States, East Asia and Japan. Microsoft reported some of the problems starting as early as 00:52 Coordinated Universal Time on Nov. 19, which is almost 5:00 p.m. Pacific Time or almost 8:00 p.m. Eastern Time on Nov. 18.

The longest service disruption lasted 11 hours, which affected some Azure Websites users in West Europe. An eight-hour disruption affected some Azure Storage users in West Europe, while some Azure Storage users in North Europe may have had a five-hour service disruption.

Microsoft MVP Troy Hunt measured an Azure disruption in the West U.S. region of more than three hours:

The partial service disruptions also may have affected the users of various Azure services, not just Websites and Storage. Here is how Microsoft described the affected services:

  • West U.S. and West Europe: Connectivity issues over five hours for some users of "Storage, Virtual Machines, SQL Geo-Restore, SQL Import/export, Websites, Azure Search, Azure Cache, Management Portal, Service Bus, Event Hubs, Visual Studio, Machine Learning, HDInsights, Automation, Virtual Network, Stream Analytics, Active Directory, StorSimple and Azure Backup Services."

  • Central U.S., East U.S., North Central U.S. and South Central U.S.: Connectivity issues over 4.5 hours for some users of "Storage, Virtual Machines, SQL Geo-Restore, SQL Import/export, Websites, Azure Search, Azure Cache, Management Portal, Service Bus, Event Hubs, Visual Studio, Machine Learning, HDInsights, Automation, Virtual Network, Stream Analytics, Active Directory, StorSimple and Azure Backup Services."

  • East U.S. 2, South East Asia and East Asia: Connectivity issues over four hours for some users of "Storage, Virtual Machines, SQL Geo-Restore, SQL Import/export, Websites, Azure Search, Azure Cache, Management Portal, Service Bus, Event Hubs, Visual Studio, Machine Learning, HDInsights, Automation, Virtual Network, Stream Analytics, Active Directory, StorSimple and Azure Backup Services."

  • North Europe, Japan East and Japan West: Connectivity issues over four hours for some users of "Storage, Virtual Machines, SQL Geo-Restore, SQL Import/export, Websites, Azure Search, Azure Cache, Management Portal, Service Bus, Event Hubs, Visual Studio, Machine Learning, HDInsights, Automation, Virtual Network, Stream Analytics, Active Directory, StorSimple and Azure Backup Services."

Possibly, this partial outage of Azure services may be Microsoft's broadest service failure to date. Microsoft had a worldwide service outage last year that also affected Azure Storage users. Nearly two years ago about 1.8 percent of Azure Storage customers experienced service outages, according to Microsoft. In the latter case, customer dashboards didn't indicate those problems for about 1.5 hours.

Microsoft Azure services typically come with a 99.9 percent service uptime service level agreement (SLA), which provides credits to customers based on any monthly downtime experienced. However, customers are responsible for actively reporting the service downtime to Microsoft in order to get any credits. A 99.9 percent SLA amounts to about eight hours of downtime per year.

About the Author

Kurt Mackie is senior news producer for the 1105 Enterprise Computing Group.