Serverless functions are a cloud computing paradigm that reduces operational overheads for developers, because the cloud provider takes care of resource management tasks such as resource provisioning, deployment, and auto-scaling. The only resource management task that developers are still in charge of is resource sizing, that is, selecting how much resources are allocated to each worker instance. However, due to the challenging nature of resource sizing, developers often neglect it despite its significant cost and performance benefits. Existing approaches aiming to automate serverless functions resource sizing require dedicated performance tests, which are time consuming to implement and maintain. In this paper, we introduce Sizeless -- an approach to predict the optimal resource size of a serverless function using monitoring data from a single resource size. As our approach requires only production monitoring data, developers no longer need to implement and maintain representative performance tests. Furthermore, it enables cloud providers, which cannot engage in testing the performance of user functions, to implement resource sizing on a platform level and automate the last resource management task associated with serverless functions. In our evaluation, Sizeless was able to predict the execution time of the serverless functions of a realistic server-less application with a median prediction accuracy of 93.1%. Using Sizeless to optimize the memory size of this application results in a speedup of 16.7% while simultaneously decreasing costs by 2.5%.
View on arXiv