Methodology or suggestion for Azure Stack SLA calculation
There is ongoing ask from Service Providers to equip them with methodology or guide how to calculate aggregated SLA for Azure Stack installations which these Service Providers are going to operate. It is obvious that there is no direct SLA commitment from Microsoft side for Azure Stack availability thus at least recommendations would be really helpful. Especially having in mind that Microsoft itself has tremendous experience to calculate SLA for its Azure DCs and Cloud Services – which hopefully can be somehow shared with Partner Ecosystem.
Thank you for your feedback.
This is currently under review
Satoshi Hara commented
My customer is also asking SLA related reference, Azure Stack. They are a service provider and they will provide IaaS service on Azure Stack to the public sector. According to them, they need to define SLA before providing service to the public sector.
I know it is very tough work to test for SLA but I would like to have some reference that meets customer's requirement
Yu Ling commented
This is also many of my customers are asking about but we can hardly provide any reliable commitment without backup from MS.
Really need an official guide as evidence.
This is great feedback.
Please continue to share your use cases
How do you define your existing SLAs and what do yo expect from this feature?
That's really huge challenge, because most of the customers want to see availability SLA of entire service in documents/treaties; But to calculate it - we need Microsoft to declare/commit AzS rack and internals availability SLA;
So, we, as CSP, can guarantee only our DC availability SLA for our customers in case of AzS now - it's not acceptable in most cases;
It would be also awesome to see some reliability for HW components (MTBF, Failure rate) from AzS Hardware vendors