Trends in Cloud-Native Performance & Efficiency

“Cloud native” technologies have existed for less than 10 years but in that short time have radically changed everything about how we define, schedule, and run workloads in the Cloud. The past few years have been marked by significant innovation in Cloud technologies— from the processors at the bottom of the stack all the way up through the software at the top of the stack. And yet, we are still catching up to the implications of all of this change for the performance and efficiency of running workloads at scale.

In this talk, we will cover: – Hardware trends in the Cloud, from multi-arch and processor price/performance analysis, to capacity, scheduling and the advent of preemptible VMs – Workload scheduling topics, such as cluster autoscaling and binpacking – OS & Kernel trends, such as the rise of the cloud OS and distroless – Changes in the container space, such as dockerless builds and multi-arch builds – Recent improvements in the JVM for performance and benchmarking – K8s interactions such as workload autoscaling, auto-tuning, and Java settings – Cloud-native developments in observability, profiling, and tooling – with real cross-cutting performance examples!