When I wrote the code to include workload identity authentication I left some edge cases unaccounted for

The code here does the following:

• a fix on the patch that checks workload identity variables while grabbing the cloud config from a secret
• lets us pass an 'Authentication' parameter in our storage classes to dictate how we want to authenticate with our storage blob (a bit more details here)
• adds logic to be selective about what annotations are added to the pvc depending on the type of authentication requested
• tries to add some hardening for when the volume attempts to be provisioned, sometimes the same pv/pvc pair runs through our block in 'nodeserver.go'

Here is a PR in hydra that uses these changes: https://msazure.visualstudio.com/One/_git/Hydra/pullrequest/8348654