Microservice architectures provide small services that may be deployed and scaled independently of each other, and may employ different middleware stacks for their implementation. Microservice architectures intend to overcome the shortcomings of monolithic architectures where all of the application's logic and data are managed in one deployable unit. We present how the properties of microservice architectures facilitate scalability, agility and reliability at otto.de, which is one of the biggest European e-commerce platforms. In particular, we discuss vertical decomposition into self contained systems and appropriate granularity of microservices as well as coupling, integration, scalability and monitoring of microservices at otto.de. While increasing agility to more than 500 live deployments per week, high reliability is achieved by means of automated quality assurance with continuous integration and deployment.

Microservice Architectures for Scalability, Agility and Reliability in E-Commerce (presentation)

5G will provide a flexible and programmable infrastructure, allowing different networks to share the same access network. A way to respond to the diverse service requirements of 5G while reducing both CAPEX and OPEX is to adopt cloud-native architectures. In this context, micro-services software design, the corner stone of cloud-native architecture, seems to be ideal for 5G. However, despite its several advantages, micro-services raise new challenges which slow its adoption down in the NFV ecosystem. Indeed, steering the expected 5G traffic between cloud-native network function is extremely challenging and is still under-investigated. In this paper, we address the service function chaining (SFC) in micro-service based network function virtualization (NFV) ecosystem from the view of the traffic steering. Specifically, we design and implement a cloud-native SFC framework offering efficient traffic steering mechanisms while considering the network state of the underlying NFV infrastructure. In this context, an optimized network-aware load balancing strategy is proposed. Based on extensive experiments, the results obtained show that our strategy achieved good results in terms of i) end-to-end latency and ii) deployment time.

Cloud-native Service Function Chaining for 5G based on Network Service Mesh

Software engineers need relevant and up-to-date architectural knowledge (AK), in order to make well-founded design decisions. However, finding such AK is quite challenging. One pragmatic approach is to search for AK on the web using traditional search engines (e.g. Google); this is common practice among software engineers. Still, we know very little about what AK is retrieved, from where, and how useful it is. In this paper, we conduct an empirical study with 53 software engineers, who used Google to make design decisions using the AttributeDriven-Design method. Based on how the subjects assessed the nature and relevance of the retrieved results, we determined how effective web search engines are to find relevant architectural information. Moreover, we identified the different sources of AK on the web and their associated AK concepts.

Exploring Web Search Engines to Find Architectural Knowledge

Context: Code and design smells, such as the coupling smells examined in this article, are widely studied. Existing empirical studies reveal gaps between the scientific theory and practice, not yet explained by the scientific literature. Only basic coupling smell detection approaches and metrics seem to have been transferred to practice so far. Objective: This article aims to study the current practitioner’s understanding of coupling smells. Method: Based on grey literature sources containing practitioner views on coupling smells, we performed a Grounded Theory (GT) study. We used UML-based modeling to precisely encode our findings and performed a rigorous analysis of our codes and models. Results: Our results are defining factors of coupling smells, as well as smell impacts, trade-offs, relationships to other smells, relationships to practices and patterns, and fix options as perceived by practitioners. We further identified gaps in the understanding of coupling smells between science and practice, and derived opportunities and challenges for future scientific work. Conclusions: Five lessons are presented as opportunities and challenges for future research. Our results can help scientists to get a better understanding of practitioner concerns, and practitioners to get an overview of the current perception of other practitioners on coupling smells.

On the practitioners’ understanding of coupling smells — A grey literature based Grounded-Theory study

A microservice architecture features hundreds or even thousands of small loosely coupled services with multiple instances. Because microservice performance depends on many factors including the workload, inter-service traffic management is complex in such dynamic environments. Service meshes aim to handle this complexity and to facilitate management, observability, and communication between microservices. Service meshes provide various traffic management policies such as circuit breaking and retry mechanisms, which are claimed to protect microservices against overload and increase the robustness of communication between microservices. However, there have been no systematic studies on the effects of these mechanisms on microservice performance and robustness. Furthermore, the exact impact of various tuning parameters for circuit breaking and retries are poorly understood. This work presents a large set of experiments conducted to investigate these issues using a representative microservice benchmark in a Kubernetes testbed with the widely used Istio service mesh. Our experiments reveal effective configurations of circuit breakers and retries. The findings presented will be useful to engineers seeking to configure service meshes more systematically and also open up new areas of research for academics in the area of service meshes for (autonomic) microservice resource management.

/pdf/an-empirical-study-of-service-mesh-traffic-management-2ohxo70l.pdf

An Empirical Study of Service Mesh Traffic Management Policies for Microservices

Microservices are becoming the de-facto standard way for software development in the cloud and in service-oriented computing. Service meshes have been introduced as a dedicated infrastructure for managing a network of containerized microservices, in order to cope with the complexity, manageability, and interoperability challenges in especially large-scale microservice architectures. Unfortunately so far no dedicated architecture guidance for designing microservices and choosing among technology options in a service mesh exist. As a result, there is a substantial uncertainty in designing and using microservices in a service mesh environment today. To alleviate this problem, we have performed a model-based qualitative in-depth study of existing practices in this field in which we have systematically and in-depth studied 40 reports of established practices from practitioners. In our study we modeled our findings in a rigorously specified reusable architectural decision model, in which we identified 14 architectural design decisions with 47 decision outcomes and 77 decision drivers in total. We estimated the uncertainty in the resulting design space with and without use of our model, and found that a substantial uncertainty reduction can be potentially achieved by applying our model.

Guiding Architectural Decision Making on Service Mesh Based Microservice Architectures

The widespread adoption of Microservice architectures has posed many challenges regarding API design for these architectures. Several API best practices and patterns have been proposed that could help API designers ensure API quality attributes such as reliability, availability, and performance. API Request Bundling, which is in focus of this paper, is one of those patterns that aims at optimizing performance. The pattern promises substantial performance gains, but can also lead to significant drawbacks such as increased development effort and application complexity. So far, there is little to no rigorously acquired knowledge to judge whether applying Request Bundling is worth the costs in a given Microservice architecture. To improve this situation, we performed an empirical study based on a Microservice-based open source business application using realistic workload scenarios. To estimate the performance impact of Request Bundling, we derived a regression model and performed a multivariate regression analysis. These selected regression models can help distributed system engineers and architects to estimate the gain in performance, in terms of round-trip time, with or without Request Bundling. Our approach followed in the paper, can be customized to other Microservice architectures or to study other performance-related Microservice patterns.

Evaluation of API Request Bundling and its Impact on Performance of Microservice Architectures

DevOps and Continuous Delivery (CD) are becoming the de-facto standard for software deployment in the cloud. Deployment pipelines are a core artefact in such practices, but so far their design is largely discussed informally, and existing sources on DevOps and CD practices are often inconsistent or incomplete. The lack of a generic, complete, tool-agnostic, and application-independent treatment of deployment pipeline design and architecture impedes their understanding by human designers and the creation of generic tools that work across different technologies and applications. To alleviate this problem, we have performed a qualitative, in-depth study of 25 deployment practice descriptions by practitioners containing informal deployment pipeline models. From our study we derived a precisely specified model of deployment pipeline architectures. We can show that the formal model substantially increases the precision of the modelling compared to the informally modelled pipelines in the original sources.

On the Design and Architecture of Deployment Pipelines in Cloud- and Service-Based Computing - A Model-Based Qualitative Study

Many API patterns and best practices have been developed around microservices-based architectures, such as Rate Limiting and Circuit Breaking, to increase quality properties such as reliability, availability, scalability, and performance. Even though estimates on such properties would be beneficial, especially during the early design of such architectures, the real impact of the patterns on these properties has not been rigorously studied yet. This paper focuses on API Rate Limit and its impact on reliability properties from the perspective of API clients. We present an analytical model that considers specific workload configurations and predefined rate limits and then accurately predicts the success and failure rates of the back-end services. The model also presents a method for adaptively fine-tuning rate limits. We performed two extensive data experiments to validate the model and measured Rate Limiting impacts, firstly on a private cloud to minimize latency and other biases, and secondly on the Google Cloud Platform to test our model in a realistic cloud environment. In both experiments, we observed a low percentage of prediction errors. Thus, we conclude that our model can provide distributed system engineers and architects with insights into an acceptable value for the rate limits to choose for a given workload. Very few works empirically studied the impact of Rate Limit or similar API-related patterns on reliability.

Amine El Malki

Papers

Guiding Architectural Decision Making on Service Mesh Based Microservice Architectures

Evaluation of API Request Bundling and its Impact on Performance of Microservice Architectures

On the Design and Architecture of Deployment Pipelines in Cloud- and Service-Based Computing - A Model-Based Qualitative Study

Impact of API Rate Limit on Reliability of Microservices-Based Architectures