A Framework for Improving Cold Start Time in Function-as-a-service (FaaS)

Abstract

This article describes a framework proposal for the cold start problem in Function-as-a-service (FaaS). The proposed framework has the goal to reduce the execution time and is presented as a prototype implemented that was evaluated with two different experimental scenarios and compared with a commercial proposal, the FaaS AWS Lambda from Amazon. The results show that the proposed framework may be considered a solution for the cold start problem and may improve the performance for applications that require a low response time.