Abstract
Recently serverless cloud computing which was proposed by Amazon in 2015 is getting more favored by developers because of its charging model of pay-by-usage, applicability for fine-grained services and transparency of the servers to developers. Although a lot of study has proved that the performance and cost of serverless platform is more ideal than the traditional cloud computing service, the new metering model of serverless which depends on execution time, usage count and memory footprint is still unfamiliar to most developers and researchers. In this paper, we firstly make an introduction to the new metering model and the advantages of serverless over traditional cloud services. Then according to the cost affecting factors, we make a comprehensive survey of some constraints that mainly impact the cost of serverless and related methods to reduce cost in three levels—function level, container level and cloud platform level. Based on the analysis, we also propose some directions worthy of in-depth study in the future.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
