Over the years, many solutions have emerged for performing service placement decisions in cloud-edge systems to optimize response time, energy usage, or cost. However, most of the solutions are theoretically evaluated or by using simulators. To tackle this issue, we conceptually designed and implemented PULCEO (Platform for Universal and Lightweight Cloud-Edge Orchestration) to contribute a platform for holistically creating, operating, monitoring, evaluating, and documenting cloud-edge orchestration environments and experimental evaluations. We used our platform to realize a representative sample topology that integrates on-premises and cloud resources. We also obtained monitoring data and performed service placement operations to analyze the application response time as an essential metric in such environments. The evaluation shows that our platform can fulfill essential requirements for cloud-edge orchestration, especially creating topologies with configurable monitoring and support for dynamically obtaining link quality metrics.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

API-Driven Cloud-Edge Orchestration with PULCEO: A Proof of Concept

  • Sebastian Böhm,
  • Guido Wirtz

摘要

Over the years, many solutions have emerged for performing service placement decisions in cloud-edge systems to optimize response time, energy usage, or cost. However, most of the solutions are theoretically evaluated or by using simulators. To tackle this issue, we conceptually designed and implemented PULCEO (Platform for Universal and Lightweight Cloud-Edge Orchestration) to contribute a platform for holistically creating, operating, monitoring, evaluating, and documenting cloud-edge orchestration environments and experimental evaluations. We used our platform to realize a representative sample topology that integrates on-premises and cloud resources. We also obtained monitoring data and performed service placement operations to analyze the application response time as an essential metric in such environments. The evaluation shows that our platform can fulfill essential requirements for cloud-edge orchestration, especially creating topologies with configurable monitoring and support for dynamically obtaining link quality metrics.