• QUIC
• FaaS
• Serverless
• Edge Computing

Serverless computing is a new trend in cloud computing technologies which involves the development of applications in the form of chaining of function calls (a.k.a. Function-as-a-Service), bare of their own internal state and performed remotely on highly scalable computing infrastructure. This stateless processes can be easily migrated or duplicated to maintain an high quality of service but one of the negative aspects of this technology concerns the application latency, which often has medium-high values and a very high variability. These aspects make serverless computing unattractive for applications adversely affected by high/variable latency.
To limit latency value/variability, the community is evaluating to run the functions on platforms closer to users, as part of the "edge computing". An unexplored possibility to take full advantages of the combined use of serverless and edge computing is the replacement of HTTP-based REST interfaces (the de-facto standard in serverless platforms) with the QUIC protocol, initially proposed by Google and today standardized by the IETF.
The aim of this thesis is to carry on an analysis of the use of QUIC as a replacement for TCP in serverless applications with very low latency. This analysis is based on the creation of a prototype developed by means of an open source library (i.e., Facebook Proxygen) and on a subsequent performance evaluation of it in a serverless edge computing representative scenario.