Kubernetes Cluster Performance Measurement on KVM-Based Nested Virtualization

Abstract

Kubernetes is an open-source container orchestration system that can be implemented in both physical and visualized infrastructure. In the implementation of virtualization, there are remaining resources that are not utilized optimally. One way to solve this problem is to implement Nested Virtualization. This study measures performance to analyze the resource usage (CPU, memory, disk, and network) of the Kubernetes cluster on KVM-based nested virtualization technology. The results of using standby and busy Kubernetes cluster memory in nested virtualization are higher, reaching 42% compared to only the Kubernetes cluster, which is 30%. CPU usage in standby and busy conditions in a Kubernetes cluster is higher at 65% compared to a Kubernetes cluster in Nested Virtualization, which is 36%. The high latency result is 237 ms on a Kubernetes cluster compared to a Kubernetes cluster on Nested Virtualization which is 40 ms. The disk read and write speeds of Kubernetes cluster technology are 410 MB/sec and 397 Mb/sec, while the Kubernetes cluster in Nested Virtualization is 116 MB/sec and 597 MB/sec. The network speed for downloading and uploading in a Kubernetes cluster is higher, namely, 39.1 Mbit/s and 53/94 Mbit/s, while the Kubernetes cluster in Nested Virtualization is 35.54 Mbit/s and 53.58 Mbit/s.