Benchmarking Data Management Systems for Microservices

Rodrigo Laigner,Yongluan Zhou
2024-05-19
Abstract:Microservice architectures are a popular choice for deploying large-scale data-intensive applications. This architectural style allows microservice practitioners to achieve requirements related to loose coupling, fault contention, workload isolation, higher data availability, scalability, and independent schema evolution.
Databases,Software Engineering
What problem does this paper attempt to address?
The problem this paper attempts to address is that existing microservice benchmarks fail to adequately reflect the core data management challenges encountered in real-world applications. This shortcoming makes it difficult to advance the development of data system technologies that support real-world, data-intensive microservice applications. Specifically, the paper points out that current microservice benchmarks are lacking in areas such as distributed transaction processing, data replication, consistent data and event querying and processing, and enforcing data integrity constraints. To tackle these challenges, the authors propose a microservice benchmark called "Online Marketplace," which aims to cover the aforementioned data management issues that have not been sufficiently addressed. Through this benchmark, researchers hope to better understand and evaluate the performance of different data management systems in handling microservice architectures, thereby driving technological advancements in this field. Additionally, the paper discusses the issues and challenges encountered when implementing different versions of the "Online Marketplace" benchmark on two different data platforms (Microsoft Orleans and Apache Statefun), and proposes a customized solution to meet all specified data management requirements.