I've been assigned the task of upgrading Terraform and the AzureRM provider . The current setup manages various Azure resources using Azure DevOps pipelines, with the Terraform backend state stored remotely in an Azure Storage Account.
Current Setup: • Terraform Version: 1.0.3 (outdated) • AzureRM Provider Version: 3.20 • Each folder represents different areas of infrastructure. Also each folder has its own pipeline. • Five Levels (Directories): • Level 1: Management • Level 2: Subscriptions • Level 3: Networking • Level 4: Security • Level 5: Compute • All levels share the same backend remote state file. • No development environment resembling production to test changes. Questions & Concerns:
- Has anyone encountered a similar upgrade scenario?
- Would upgrading AzureRM from 3.20 to 3.117 modify the state file structure?
- If we upgrade one level at a time (e.g., Level 1 first, then Level 2, etc.), updating resource blocks as needed, will the remaining levels on 3.20 continue functioning correctly until they are also upgraded? Or could this create compatibility issues? I haven’t made any changes yet and would appreciate any guidance or best practices before proceeding. Looking forward to your insights!
