Why ServerlessBench?

• Serverless-customized metrics

  ServerlessBench evaluates various metrics in different aspects of Serverless computing, including performance, composition methods, and so on. It helps users to thoroughly profile and analyze serverless applications.

• Diverse function languages

  ServerlessBench consists of functions writen with multiple languages, including C, Java, Python, NodeJS and Ruby. It enables the analysis of the characteristics of Serverless applications written by different languages. Moreover, it is convenient for different language developers to use.

• Supporting different serverless platforms

  ServerlessBench comprises of workloads designed for open-source platforms (OpenWhisk, Fn) and popular commercial public cloud (AWS Lambda). Users can not only evaluate the complete process of a serverless application from serverless platform developers perspective, but also characterize serverless applications on commercial platforms from serverless computing users perspective.

• Real-world applications

  ServerlessBench collects a set of popular real-world applications, representing the features of serverless applications in the real world.