Pulumi provider doesn't work without `jwtRbac: true`

[fredtcaroli]

First, I'm not sure if this is the right place to post Pulumi-related issues, but I assume this would also happen with Terraform, I just don't have the setup to test it.

The Issue

I'm at 1.15.0 and when I try to include something like this to my Pulumi stack:

    const qdrantCluster = new qdrantCloud.AccountsCluster('qdrant-cluster', {
      cloudProvider: 'aws',
      cloudRegion: region.name,
      configuration: {
        numberOfNodes: 1,
        nodeConfiguration: {
          packageId: packages[0]!.id,
        },
      },
    });

It deploys fine at first, but then if I do another pulumi up without any changes the terraform provider will fist detect some spurious changes, and then fail completely to deploy with a very cryptic message:

  qdrant-cloud:index:AccountsCluster (qdrant-cluster):
    error: error updating cluster [6e218129-97c5-49ba-9c94-ddcb1b241103|6e218129-97c5-49ba-9c94-ddcb1b241103]: rpc error: code = Internal desc = Internal server error: 

After hours of debugging, I realized the examples in the documentation all included the jwtRbac option, so I decided to try it:

    const qdrantCluster = new qdrantCloud.AccountsCluster('qdrant-cluster', {
      cloudProvider: 'aws',
      cloudRegion: region.name,
      configuration: {
        numberOfNodes: 1,
        databaseConfiguration: {
          service: {
            jwtRbac: true,
          },
        },
        nodeConfiguration: {
          packageId: packages[0]!.id,
        },
      },
    });

Now the spurious diff continues to happen, but at least the provider is successful at not barfing the deployment.

Expected Behavior

1. The deployment should not fail without the jwtRbac OR the jwtRbac should be required.
2. The terraform provider should not be detecting changes where there are none.

[bashofmann]

Thank you for your report. We can validate the issue and are working on a fix.