After reading this article I would offer up some food for thought from each of the 5 points…
“It’s time to sit down with architects to map out the ideal path to realizing the benefits of becoming truly cloud-native”.
I would challenge that it is NOT about becoming cloud-native. It is better to think along the lines what applications can take advantage of being containerized and which cannot, or in fact should not since in many cases “it’s not broken so don’t fix it”. You can then think about where best to run those containerized workloads, including on the recaptured compute potential of the legacy servers for those applications that continue to run as-is. The best of both worlds.
“What is chosen today may not necessarily make sense for the business years from now”.
We live in a world where everything is cyclical. Just think how we started (and I date myself here) with centralized compute power and storage in mainframes connected to users via dumb green-screen terminals. We then evolved to client-server as the compute power at the user end grew and we pushed out from a centralized paradigm to take advantage of this power. The pinnacle reached with desktop grids connecting all these compute resources together to provide the equivalent of a super-computer. The world turned again and the (public) cloud shiny object was thrown out there so all the compute power again pushed to become centralized, but now in someone else’s data center, connected via dumb browsers. Now we turn back to private cloud as the publicized benefits of public cloud aren’t all they are cracked up to be. Hybrid cloud is the mantra in 2018. Perhaps with desktops and other compute devices (including IoT) still continuing to increase in power we will see a new “desktop grid” in 2019 and beyond?
“Leverage existing infrastructure and the investment made in on-premise data centers”.
Absolutely. To paraphrase from my first point. Put the right workload, on the right resource, for the right reasons (whether it be cost, security, compliance, network, or whatever).
“Modernizing an existing data center environment gives enterprises more runway to test and incrementally validate their cloud migration strategy.”
More than this, modernizing the data center provides increased opportunity, flexibility, and more importantly CONTROL, not to mention that as servers (and desktops) continue to refresh with increased compute and storage capabilities they represent even more on-going potential as recaptured idle/unused resource.
“If an organization is too focused on the traditional data center or cloud, these edge experiences can suffer.”
Yes. Organizations need to consider the entire compute continuum, from desktops all the way out to the Edge and IoT, and perhaps the idea of a next-generation compute-device grid could be around the corner.
And to bring it all together my final thought is although “Winter is coming for enterprise cloud” thinking outside the pundits’ cloud box to more expansive, creative, IT strategies will lead to the, yet unrealized, possibilities of Spring.