The whole time I’ve been a cloud and AI technical trainer, it has been funded somehow by a given hyperscale cloud provider. I always get interesting questions from my students about various multi-cloud scenarios, and I’m always interested in learning more and helping customers learn more – but the training content never accounts for the really popular multicloud scenarios.
The recent big AWS outage kinda shows why, ya know? Because the multi-cloud thing real customers are the MOST interested in is hedging their bets.*
- Can we Terraform deployments to all of the big 3 (western) clouds, so we can still have our mission-critical stuff up and running in the event of a major outage disaster?
- Can we replicate important content to buckets everywhere, to second guess the purported reliability of any one of these big three players?
I get it – no major cloud provider is incentivised to acknowledge the possibility of their own disasters, or create training materials on how to migrate stuff smoothly to their competitors.
*Notably, I don’t usually get excited questions about hybrid – customers who are all in on cloud don’t want to have to manage a datacenter themselves. They just want to distribute things out as much as possible, just in case.
This post appeared initially on LinkedIn; I finally migrated it here.