Optimal and Automated Deployment for Microservices

Abstract

Microservices are highly modular and scalable Service Oriented Architectures. They underpin automated deployment practices like Continuous Deployment and Autoscaling. In this paper we formalize these practices and show that automated deployment -proven undecidable in the general case -is algorithmically treatable for microservices.Our key assumption is that the configuration life-cycle of a microservice is split in two phases: (i) creation, which entails establishing initial connections with already available microservices, and (ii) subsequent binding/unbinding with other microservices. To illustrate the applicability of our approach, we implement an automatic optimal deployment tool and compute deployment plans for a realistic microservice architecture, modeled in the Abstract Behavioral Specification (ABS) language.

Operating Latency Sensitive Applications on Public Serverless Edge Cloud Platforms

“…Researchers in [11], [21] study the optimization problem of cloud native service composition and provide offline solutions based on game theoretic formulation and a constrained shortest path problem. Other recent works in [5], [8], [22] target similar problems of performance optimization of serverless applications leveraging public cloud resources, but only regarding the placement problem of the service components and missing any adaptive and automated service reoptimization task.…”

A Formal Model of the Kubernetes Container Framework

“…The ABS code emulating the clients can be found in the repository of the simulator along with the Jmeter stress test descriptor. 9 Experiment 2. The purpose of Experiment 2 is to test the prediction ability of the modeling framework.…”

“…It has been shown that deployment management can be formalized as finite state machines, such as the Aeolus [13] and TOSCA-compliant deployment models [10], which can be adapted to formally reason about the static deployment of microservices; i.e., to express component resilience and static links between components. For example, the static deployment of microservices can be encoded as a constraint problem [9]. This work, which is based on Aeolus, takes an ABS model as its starting point.…”

Solver-Aided Multi-Party Configuration

“…General research on configuration synthesis has thrived, yielding, over the past decades, tools to produce configurations for firewalls (e.g., [12,13,17,42,54]), routing (e.g., [4,5,15]), and other networking targets [14,38,44,47]. Some work, such as Bravetti, et al [7] also focuses on microservice synthesis. All of these generate a monolithic configuration, sidestepping the multi-party case.…”