Tetris: An SLA-aware Application Placement Strategy in the Edge-Cloud Continuum

TL;DR

This work tackles SLA-driven modular application placement in the edge-cloud continuum by introducing Tetris, a heuristic that tightly couples scheduling and placement through a composite SLA metric $\phi(t)$ and a capacity-balancing metric $\gamma(v)$. By prioritizing tasks with urgent deadlines and distributing them to nodes with balanced resources, Tetris reduces SLA violations and prevents task drops, achieving about a 76.5% improvement over the state-of-the-art Thea across a 7-node testbed. The approach also analyzes the impact of cloud availability on latency and energy, showing that edge-cloud configurations improve SLA compliance at the cost of higher average latency and modestly higher power. Overall, Tetris delivers near-optimal SLA adherence and resilience under dynamic workloads by integrating latency, deadline, drop penalties, and energy considerations into a unified placement framework.

Abstract

An Edge-Cloud Continuum integrates edge and cloud resources to provide a flexible and scalable infrastructure. This paradigm can minimize latency by processing data closer to the source at the edge while leveraging the vast computational power of the cloud for more intensive tasks. In this context, module application placement requires strategic allocation plans that align user demands with infrastructure constraints, aiming for efficient resource use. Therefore, we propose Tetris, an application placement strategy that utilizes a heuristic algorithm to distribute computational services across edge and cloud resources efficiently. Tetris prioritizes services based on SLA urgencies and resource efficiency to avoid system overloading. Our results demonstrate that Tetris reduces SLA violations by approximately 76% compared to the baseline method, which serves as a reference point for benchmarking performance in this scenario. Therefore, Tetris offers an effective placement approach for managing latency-sensitive applications in Edge-Cloud Continuum environments, enhancing Quality of Service (QoS) for users.

Figures (13)

• Figure 1: Edge–Cloud Continuum Environment
• Figure 2: Overview of Tetris algorithm concept
• Figure 3: Allocation Request
• Figure 4: Toy example of Thea algorithm
• Figure 5: Toy example of Tetris algorithm