Are you heavily considering a multi cloud approach after AWS’ last week US-East region catastrophic outage? That may be the correct answer for some of your applications that were affected, but it is not as simple as it sounds.
To properly deliver multi-cloud resiliency it takes proper understanding, planning, testing, and much more! And let’s face it, there will always be a single point of failure somewhere along the way (DNS, load balancer VIP, network route, MFA, etc…).
Why jump straight into a multi cloud conversation after this incident? Not saying that it is wrong to have certain workloads running in multiple clouds, however, you could have been leveraging multiple regions within AWS to begin with and you could have avoided some headaches.
Also, what about cost? How much revenue does this application produce for the business before you go on a shopping spree to deploy the same type of instances in multiple AZ/regions/clouds? Are you able to justify spending 4x more in order to possibly reduce the downtime in the future?
I still believe, even after this region outage that you should first guarantee that high resiliency within a single cloud leveraging their multiple regions, and only then consider adding another layer of complexity (another cloud provider) in order to guarantee even higher resiliency.
