Infrastructure as code and automation tools for efficient multi-cloud resource provisioning

Cloud computing, with its features such as low initialization cost, low long-term cost, and high scalability, provides technical support for enterprises to build large-scale information systems and adopt advanced technologies such as big data, artificial intelligence, and the Internet of Things. With the rapid development of cloud computing, especially in the last decade, various cloud service providers have sprung up. Enterprises are provided with a cloud-computing platform with different service modality of Infrastructure as a Service, but they have been struggling with resource management and utilization. Given the resource management and utilization become increasingly complicated with the extremely wide variety of factors involved in the decision-making process, it has made the whole process inefficient and ineffective. Workload Management Systems are commonly used by Service Providers and cloud customers to manage the distribution of workloads among resources. However, they are designed for a single-cloud environment and cannot efficiently and effectively manage the workflow tasks in a Multi-Cloud Environment (Lakshman & Malik, 2010; Sharma & Lamba, 2020; Liu & Shao, 2021).

Due to the emergence of multi-cloud systems, it is getting common for an enterprise to utilize resources from multiple cloud service providers. And for an enterprise to efficiently and effectively utilize the resources from the Cloud Service Providers, it prefers to set up a Multi-Cloud Environment which consists of various types of resources from diverse providers. However, the heterogeneous properties of the assets, works, and service levels in a Multi-Cloud Environment with multiple assets originating from multiple providers have rendered the workflow management a much more complicated problem than for a single-cloud system. On the one hand, for a more complicated multi-cloud environment with more diverse conditions, objectives, properties, and specifications of different resources, it is getting demanding and crucial that a workflow is optimized according to its defined settings and characteristics. On the other hand, with the development of cloud-computing technologies and innovations such as storage management and deployment technologies, more and more large-scale workflows are migrated to the cloud for management and execution. It has consequently increased the pressure on the interest convergence of multiple providers in the booming multi-cloud system (Zaharia et al., 2010; Tsai et al., 2019).