Microsoft’s Azure cloud computing platform suffered a series of outages and service disruptions in the past week that affected several products and impacted customers in various parts of the world.
Maintaining solid performance and reliability of the Azure infrastructure- and platform-as-a-service cloud tools is key for Microsoft, as it battles strong rivals in this market, like Amazon and Google.
Microsoft CEO Satya Nadella has identified the Azure suite of PaaS and IaaS products as essential for the current and future financial success of the company, which is shifting its focus from providing software that customers install on their premises to selling cloud services and applications hosted on its data centers.
Microsoft didn’t immediately respond to a request for comment regarding the incidents that affected Azure in the previous seven days, as logged in this Azure service status page.
On Friday of last week, Azure customers in Japan suffered what Microsoft described as a “partial performance degradation,” which wasn’t fully resolved until Monday. It included auto-scaling problems with Cloud Services, designed for customers to build, deploy and manage apps on the Azure cloud, and with Virtual Machines, which lets customers deploy Windows Server, Linux, or third-party software images to Azure.
Auto-scaling is an Azure feature that lets customers configure Azure to automatically and dynamically adjust the performance of their application based on pre-set metrics and parameters.
“Additionally, customers may have seen issues with their Metrics Reporting Services within their Management Portal,” the Azure status log reads.
Also last Friday, a couple of “partial service interruption” incidents in the western U.S. region affected a subset of Cloud Services, Storage and Backup services customers.
Then for about four hours on Monday evening U.S. Eastern Time, “intermittent connectivity issues” affected Azure services deployed in the southern Brazil region, an issue Microsoft described as a “partial service interruption” related to network infrastructure.
On Tuesday, Microsoft resolved a “partial performance degradation” issue that cropped up a week prior affecting customers in the eastern U.S., who “may have experienced issues with proper data flow to their Auto-Scale configurations.”
A “full service interruption” hit Visual Studio Online in multiple regions between Wednesday and Thursday. It started with performance problems on Wednesday evening that lingered through late Thursday morning. Then that was followed by a full-blown outage of Visual Studio Online that lasted about six hours.
On Thursday evening, Microsoft zapped a “partial performance degradation” issue that appeared Aug. 1, in which “a subset of customers in multiple regions may have experienced that their auto-scaling ability was reduced or disabled.”
Also on Thursday evening, for almost two hours, a “limited subset of customers” across multiple regions may have experienced issues logging into the Azure Management Portal.