Paper Details

Name: Kartik Bhardwaj, Alish Pandey, Rhythmpreet Kaur & Ramandeep Singh
Country: India

Year: 2026
Volume: Volume No: 13, January, Year: 2026 (Special Issue)
Page Number: 150 -154
DOI: https://doi.org/10.5281/zenodo.18953393

ABSTRACT
Governance and provisioning of Azure resources across various subscriptions remains a challenging task, largely due to the limitations of Terraform’s azurerm provider by design. While Terraform enjoys widespread acclaim as an Infrastructure-as-Code (IaC) capabilities fall short in tool, multi-subscription automation. It describes a novel approach combining python scripting, pipeline automation using YAML in azure devops and Terraform to achieve a scalable, performant and fully automated resource provisioning spanning multiple Azure subscriptions. In particular, Terraform variable files ( .tfvars ) with subscription data imported from a structured CSV file, eliminating the need for any manual configuration. A matrix strategy on a YAML pipeline orchestrates the run of Terraform steps—init, plan and apply— in parallel across multiple workspaces significantly speeding up deployment while reducing overhead. For large-scale, enterprise deployments, this setup is particularly useful as storing Terraform state files in a centralized Azure Storage account enhances both maintainability and conflict prevention. So presenting a systematized automation-based method that minimizes human interaction and streamlines the provisioning process while improving consistency of state is the significant finding of this work, translating into a better cloud automation approach. Well suited for multi-cloud deployments, the approach is both a sound architectural and operational best practice that help alleviate some of the most common scalability and operational challenges that all cloud practitioners experience.

Keywords: Azure, Terraform, Multi-Subscription Automation, YAML Pipelines, Python Scripting, Cloud Automation.