Azure goes down, this is a super big deal

Azure MFA (multi-factor authentication) failed again last week, the second major outage in a month following the 17-hour failure which impacted customers in Europe. This is a *really* big deal. For readers who may not know how Azure MFA works, basically all Azure services which require second factor authentication (e.g. a code by text, or an automated phone based security, security key, etc.) rely on Azure MFA, and customers are locked out of their Microsoft services when MFA does not work properly.

The cause of this most recent outage was apparently related to DNS. This news made me think of a judo buddy who worked as an engineer for a Microsoft cloud based search service when it originally started, and how they would be on call via pager (yes, the world has changed) with 5 minute response times in case of service degradation ... I joked with him that cloud was no different than regular IT, where almost all services are vulnerable to outages because of either DNS or certificates. It's bizarre to me that Microsoft with all of its resources cannot mitigate these type of well know issues with its space age cloud services when they're known even to lowly IT consultants like me ...

I trust Microsoft's cloud more than I trust a lot of other services, including on-premise services. But as we move into 2019, it's surely the case that a customer's cloud services and process design have got to reflect the reality that these services are not yet 5 9s, and some amount of flexibility is essential. A seasoned cloud consultant is going to know the weaknesses in all types of systems and help guide projects to mitigate these types of failures. Let's hope that Microsoft understands the impact on trust that these issues create, and that they move swiftly to improve their messaging on these issues, and to address any gaps or weaknesses in their services.