GreenFaaS: Maximizing Energy Efficiency of HPC Workloads with FaaS

TL;DR

GreenFaaS is presented, a novel open source framework that bridges the gap between energy-efficient applications and FaaS platforms and demonstrates that intelligent placement of tasks can both reduce energy consumption and improve performance.

Abstract

Application energy efficiency can be improved by executing each application component on the compute element that consumes the least energy while also satisfying time constraints. In principle, the function as a service (FaaS) paradigm should simplify such optimizations by abstracting away compute location, but existing FaaS systems do not provide for user transparency over application energy consumption or task placement. Here we present GreenFaaS, a novel open source framework that bridges this gap between energy-efficient applications and FaaS platforms. GreenFaaS can be deployed by end users or providers across systems to monitor energy use, provide task-specific feedback, and schedule tasks in an energy-aware manner. We demonstrate that intelligent placement of tasks can both reduce energy consumption and improve performance. For a synthetic workload, GreenFaaS reduces the energy-delay product by 45% compared to alternatives. Furthermore, running a molecular design application through GreenFaaS can reduce energy consumption by 21% and runtime by 63% by better matching tasks with machines.

Figures (10)

• Figure 1: Per-machine runtime, energy, and average power for the Graph Pagerank benchmark.
• Figure 2: Per-function runtime, energy, and average power for eight serverless benchmark suite functions on Institutional Cluster.
• Figure 3: Runtime and energy comparison of benchmark tasks across the four systems. Each values is normalized to the average for the corresponding task across all systems.
• Figure 4: GreenFaaS high-level architecture, showing integration with Globus Compute and Transfer.
• Figure 5: Globus web app with the bookmarklet enabled.