Serverless Straggler Mitigation using Local Error-Correcting Codes

Abstract

Inexpensive cloud services, such as serverless computing, are often vulnerable to straggling nodes that increase end-to-end latency for distributed computation. We propose and implement simple yet principled approaches for straggler mitigation in serverless systems for matrix multiplication and evaluate them on several common applications from machine learning and high-performance computing. The proposed schemes are inspired by error-correcting codes and employ parallel encoding and decoding over the data stored in the cloud using serverless workers. This creates a fully distributed computing framework without using a master node to conduct encoding or decoding, which removes the computation, communication and storage bottleneck at the master. On the theory side, we establish that our proposed scheme is asymptotically optimal in terms of decoding time and provide a lower bound on the number of stragglers it can tolerate with high probability. Through extensive experiments, we show that our scheme outperforms existing schemes such as speculative execution and other coding theoretic methods by at least 25%.

Utility-based Resource Allocation and Pricing for Serverless Computing

“…Note that providing such strict and premium SLAs require high-level maintenance of the cloud, the costs of which are indirectly borne by all the users. This may be unfair since some users could be equipped to tolerate job failures [32,33] and different users could have varying degrees of fault-tolerance. An interesting future direction is to design pricing schemes that conform with each user's preferences using behavioral preference models from decision theory (see, for example, [34]).…”