Oh What a Fine Mess We’ve Made: The State of Cloud Automation
More than a decade ago I worked on a team at Intel that introduced a vision for the future of cloud based on concepts of federation and automation. The goal was to build the cloud as a self-aware infrastructure where services were automatically provisioned and the cloud itself would monitor and re-balance resources based on workload usage. A colleague later coined this as a data center that thinks for itself.
As technology has progressed, we have made fantastic strides in delivering the core capabilities for this vision. Cloud native computing with the workload portability required for re-provisioning. Underlyling infrastructure that has flexible performance capabilities to manage different workload types, and the promise of infrastructure composability on the horizon. And stack automation giving us the ability to manage orchestration by policy. AI promises even more advancements towards that data center that can actually think through decisions historically managed by humans.
So why haven’t we achieved this nirvana? You could argue that the cloud itself has begat the core issue. Its very nature has made it incredibly easy to spin up workloads, different types of workloads (serverless, microservices etc) and this complexity has given rise to a equally complex set of tools to oversee more complex allocation of resourcing. I got a chance to talk to Abby Kearns, a leading innovator in cloud computing about the challenge. Abby is a powerhouse when it comes to deeply understand cloud stacks. As the former CTO of Puppet, she led the app stack automation company’s delivery of technology that helped inspire broad industry innovation in the space. Her viewpoints are also shaped by her time leading the Cloud Foundry, a powerful cloud consortium.
In our discussion, Abby pointed to this complexity as a key challenge that gates IT oversight today and offers some hope on cloud stack innovation. Check out the interview and as always thanks for engaging - Allyson
More than a decade ago I worked on a team at Intel that introduced a vision for the future of cloud based on concepts of federation and automation. The goal was to build the cloud as a self-aware infrastructure where services were automatically provisioned and the cloud itself would monitor and re-balance resources based on workload usage. A colleague later coined this as a data center that thinks for itself.
As technology has progressed, we have made fantastic strides in delivering the core capabilities for this vision. Cloud native computing with the workload portability required for re-provisioning. Underlyling infrastructure that has flexible performance capabilities to manage different workload types, and the promise of infrastructure composability on the horizon. And stack automation giving us the ability to manage orchestration by policy. AI promises even more advancements towards that data center that can actually think through decisions historically managed by humans.
So why haven’t we achieved this nirvana? You could argue that the cloud itself has begat the core issue. Its very nature has made it incredibly easy to spin up workloads, different types of workloads (serverless, microservices etc) and this complexity has given rise to a equally complex set of tools to oversee more complex allocation of resourcing. I got a chance to talk to Abby Kearns, a leading innovator in cloud computing about the challenge. Abby is a powerhouse when it comes to deeply understand cloud stacks. As the former CTO of Puppet, she led the app stack automation company’s delivery of technology that helped inspire broad industry innovation in the space. Her viewpoints are also shaped by her time leading the Cloud Foundry, a powerful cloud consortium.
In our discussion, Abby pointed to this complexity as a key challenge that gates IT oversight today and offers some hope on cloud stack innovation. Check out the interview and as always thanks for engaging - Allyson
