Last week, two incident disrupted the service of 26 regions out of 28 for Microsoft Azure’s public cloud. One incident was caused by a software error and may have kept customers from during on diagnostics or using other resources. The other incident lasted about 8 hours and caused outages for many Azure services.
In 2012, Azure suffered a 12-hour outage on leap day, which caused the company to refund one-third of the bill of customers. In 2013, the company suffered a partial performance degradation and in 2014, a shorter service disruption. This isn’t anything new for the company.