Fault Tolerance in Real-Time Cloud Computing

Abstract

This paper presents the Fault-Tolerant Real-Time Cloud (FTRTC) project that aims to design cloud computing infrastructures capable of hosting highly reliable and real-time applications. These applications are characterized by strict timing and reliability constraints, as well as critical failure scenarios. For instance, such requirements are commonly found in the context of Industry 4.0. We present a formalization of the problem of designing real-time cloud applications supporting an adjustable level of fault tolerance throughout their distributed execution in a cloud infrastructure. The contributions presented in this paper indicate important research directions when building cloud infrastructures able to supporting ultra-reliable real-time applications.

Towards a Holistic Cloud System with End-to-End Performance Guarantees

“…Moreover, fault-tolerance and the related repercussions on service availability are often neglected in these works, despite their relevance in real-world cloud platforms (and for the proper functioning of time-critical applications). In this regard, I contributed to a novel performance model for time-critical, fault-tolerant cloud architectures [1], [6] in collaboration with Ericsson Research. Our work focuses on estimating at design-time the worst-case response time for a set of microservice-based time-critical applications by taking into account the worst-case workload patterns, as well as the occurrence of faults.…”

“…As of today, no cloud provider stipulates a Service-Level Agreement (SLA) for individual request guarantees, but rather for a number of requests over a wide time period (i.e., months, years). For instance, AWS DynamoDB [10] and Google BigTable [9] are two fully-managed cloud service that claim to guarantee single-digit millisecond per-request latency; nevertheless, their SLAs are defined in terms of the "usual" monthly service availability 1,2 .…”