Pipeline deployment deletes using filter mlz- prior to testing deployment which causes issues as its too generic

[shawngib]

Description

During Azure Devops deployment uses az cli to delete diagnostic settings and resource groups in pre-task to deployment as a way to solve for previous issues in deployments leaving orphaned resource blocking future deployments. While this is a temporary fix it impacts testing and should be more specific in nature. Cleanup should be after deployment and specific to name of deployment and result in failure if not completed.

current inline script:
'az group list -o table | grep ''mlz'' | awk ''{system("az group delete -y --no-wait -g "$1)}''

• Different results in Bicep versus TF.
  -- TF runs destroy
  -- Bicep doesn't clean up after deployment

Steps to Reproduce

Steps to reproduce the behavior:

1. Run pipeline manually or triggered

Expected behavior

• Success should reflect deployments in Gov and Commercial for Bicep az deployment and terraform apply with no artifacts left behind in subscription
• Cleanup script should be run (modify script to wait for results) filtering on the output resourcePrefix of deployments and pass or fail since clean up is also a requirement of good IaC.

Actual behavior

-- Bicep resources left behind, next deployment generically deletes anything with `,lz' in name prior to next

Screenshots

Additional context

Operating System:
Terraform Version:
Cloud (public, Azure Government, etc.):

[brooke-hamilton]

Closing - the automation did not close the issue along with the PR completion.