Resumen:
|
[EN] Cloud Computing allows companies to scale seamlessly, providing a broad range of state-of-the-art services. Another promise is to free users from the operational and administrative burdens [1]. However, with the advent ...[+]
[EN] Cloud Computing allows companies to scale seamlessly, providing a broad range of state-of-the-art services. Another promise is to free users from the operational and administrative burdens [1]. However, with the advent of cloud-native applications [2], this promise becomes questionable – especially when DevOps [3] principles are used during development. Experience from practice shows that teams often struggle dealing with both the infrastructure, finding the right architecture, and implementing business logic. When working in decentralized teams, things are even worse, as standardization across teams cannot be assumed. To tackle those issues, automation by means of techniques such as Infrastructure-as-code [4] help to ease to cope with infrastructural concerns. However, when working with several teams in a decentralized manner, operational overhead is still there. Organizations struggle with standardization of infrastructure code and there is no clear centralized visibility for security-related concerns within the development lifecycle. To address these issues, we propose two things: First, building up a Platform Team [5], which serves as an organizational structure for continuous delivery. A Platform Teams can be the size of a typical small DevOps Team and support the whole organization with standardized security-hardened modules. Second, an Ops-Platform is needed that is operated by the Platform Team to centrally provide and maintain those modules. Other Dev-Teams can then consume those modules. In this paper, we report insights from the implementation of this approach in practice. We find out that developers are 75% less focused on operations by using such a platform and name specific success factors. References [1] Borenstein, Nathaniel, and James Blake. "Cloud computing standards: Where's the beef?." IEEE Internet Computing 15.3 (2011): 74-78.[2] Kratzke, Nane, and Peter-Christian Quint. "Understanding cloud-native applications after 10 years of cloud computing-a systematic mapping study." Journal of Systems and Software126 (2017): 1-16.[3] Ebert, Christof, et al. "DevOps." Ieee Software 33.3 (2016): 94-100.[4] Morris, Kief. Infrastructure as code. O'Reilly Media, 2020.[5] Leite, Leonardo, et al. "Platform teams: An organizational structure for continuous delivery."Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops. 2020.
[-]
|