Swarm and Mesos are not quite ready to throw in the towel quite yet so multiple orchestration engines will persist. Kubernetes is complicated and low and behold a myriad of vendors rushing to make it simpler. And last year public cloud first gave way to hybrid/multi-cloud and with the advent of Edge, IoT, and even desktop recapture, the conversation is even now broadening beyond just “cloud” as we head into 2019. The result – multiple orchestrators, complexity, and a need to run not just “on both your own data center and private cloud and, when needed, a public cloud” but on an ever-expanding continuum of compute resources.
Solutions such as Federation are far away from becoming a practical commercial reality and even when available won’t address the multiplicity of orchestrators or the inevitable technology reality that Kubernetes will no longer be the “best thing since sliced bread” and be supplanted by the next big thing. Don’t we need one orchestrator to rule them all? One designed not just for cloud, but any and all resources, not to any single orchestrator, but any and all as circumstance dictates, and with the (AI) intelligence to deliver on what the end goal should be – Optimal Workload Placement across an organizations complete IT infrastructure, today and tomorrow.