Low-code ChatOps for Microservices Systems Using Service Composition
Shengkai Wang, Shang‐Pin Ma, C.-W. Chao, Guan-Hong Lai · 2023
The core concepts of microservices architecture, such as service independence and autonomy, small teams, and continuous evolution, align well with the principles of DevOps (development and operations). However, the distributed nature of microservices architecture introduces complexities in development and operations at various levels. Microservice teams often need to adopt multiple DevOps tools, increasing the learning curve and making it challenging for teams and stakeholders to grasp the system’s current status. Additionally, team members must gather development and operations information from multiple sources, resulting in increased effort and a lack of cohesive integration and usage. To address these issues, we propose a ChatOps mechanism specifically tailored for microservices architecture, called ChatOps4Msa (ChatOps for Microservices Architecture). Via ChatOps4Msa, users can monitor real-time service status, perform service testing, track issues and deployment logs, and receive periodic alerts and important information using a query language devised in this research, called CQL (ChatOps Query Language). ChatOps developers (CODs) only need to write configuration files using a low-code syntax proposed in this research to build the required capabilities quickly. Besides, CODs can also create new capabilities by composing services from multiple DevOps tools. The results of the experimental evaluation show that ChatOps4Msa are effective in improving the adoption and usability of the DevOps toolchain.