GeoFF: Federated Serverless Workflows with Data Pre-Fetching

2024-05-22

2024 IEEE International Conference on Cloud Engineering (IC2E)

AI Summary

To address high end-to-end latency in cross-platform Function-as-a-Service (FaaS) workflows—caused by hardcoded logic, cascading cold starts, inter-function latency, and data-download bottlenecks on critical paths—this paper proposes a federated serverless workflow orchestration middleware. The middleware enables collaborative execution across public and private FaaS platforms, introducing the first dynamic heterogeneous-platform workflow reorchestration mechanism. It integrates function prewarming and data prefetching to eliminate cold starts and data loading from critical paths. Leveraging cross-platform API abstraction and a lightweight orchestration engine, it achieves transparent, platform-agnostic scheduling. Evaluated in realistic deployments, the approach reduces end-to-end latency by over 50%, significantly surpassing conventional serverless workflow performance limits. Results validate the effectiveness and scalability of federated execution coupled with proactive prefetching optimization.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Hardware-aware ML
• Search and Optimization: Distributed Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work

Abstract

Function-as-a-Service (FaaS) is a popular cloud computing model in which applications are implemented as workflows of multiple independent functions. While cloud providers usually offer composition services for such workflows, they do not support cross-platform workflows forcing developers to hardcode the composition logic. Furthermore, FaaS workflows tend to be slow due to cascading cold starts, inter-function latency, and data download latency on the critical path. In this paper, we propose GEOFF, a serverless choreography middleware that executes FaaS workflows across different public and private FaaS platforms, including ad-hoc workflow recomposition. Furthermore, GEOFF supports function pre-warming and data pre-fetching. This minimizes end-to-end workflow latency by taking cold starts and data download latency off the critical path. In experiments with our proof-of-concept prototype and a realistic application, we were able to reduce end-to-end latency by more than 50%.

Problem

• Enables cross-platform serverless workflow execution
• Reduces latency from cold starts and data transfer
• Supports dynamic workflow recomposition and pre-fetching

Innovation

• Cross-platform serverless choreography middleware for FaaS workflows
• Function pre-warming to eliminate cold start delays
• Data pre-fetching to reduce download latency significantly