cluster.get_kubeconfig() doesn't work with pulumi preview when gcp project is created programmatically and project_id is generated

[jancespivo]

What happened?

cluster.get_kubeconfig() doesn't work with pulumi preview when gcp project is created programmatically and project_id is generated via pulumi_random

pulumi up --skip-preview works fine.

When project_id is set manually, it works fine.

Example

import pulumi
import pulumi_gcp as gcp
import pulumi_google_native as google_native
import pulumi_google_native.container.v1 as container
import pulumi_kubernetes as kubernetes
import pulumi_random as random

random_string = random.RandomString(
    "project-id-suffix",
    length=6,
    special=False,
    lower=True,
    upper=False,
    number=False,
).result

config = pulumi.Config()

gcp_project = gcp.organizations.Project(
    "cluster",
    name="minimal",
    project_id=pulumi.Output.concat("minimal-", random_string),
    billing_account=config.require("gcp_billing_account_id"),
)

gcp_provider = gcp.Provider(
    "gcp-provider",
    project=gcp_project.project_id,
)

google_native_provider = google_native.Provider(
    "google-native-provider",
    project=gcp_project.project_id,
)

compute_service = gcp.projects.Service(
    "compute.googleapis.com",
    service="compute.googleapis.com",
    opts=pulumi.ResourceOptions(
        providers={
            "gcp": gcp_provider,
        },
    ),
)

container_service = gcp.projects.Service(
    "container.googleapis.com",
    service="container.googleapis.com",
    opts=pulumi.ResourceOptions(
        depends_on=[compute_service],
        providers={
            "gcp": gcp_provider,
        },
    )
)

cluster = container.Cluster(
    "cluster",
    node_pools=[
        container.NodePoolArgs(
            name="default",
            initial_node_count=1,
            config=container.NodeConfigArgs(
                machine_type="e2-custom-4-8192",
                disk_type='pd-balanced',
            ),
            locations=[
                "europe-west1-b",
            ]
        )
    ],
    initial_cluster_version="1.25",
    location="europe-west1-b",
    opts=pulumi.ResourceOptions(
        depends_on=[container_service],
        providers={
            "google-native": google_native_provider,
        },
    )
)

k8s_provider = kubernetes.Provider(
    "k8s_provider",
    kubeconfig=cluster.get_kubeconfig(),
)

pulumi preview or pulumi up raises

error: Program failed with an unhandled exception:
    Traceback (most recent call last):
      File "/home/beer/.pulumi/bin/pulumi-language-python-exec", line 197, in <module>
        loop.run_until_complete(coro)
      File "/usr/lib/python3.11/asyncio/base_events.py", line 653, in run_until_complete
        return future.result()
               ^^^^^^^^^^^^^^^
      File "/home/beer/projects/tacr/platform/cluster-deploy/.venv/lib/python3.11/site-packages/pulumi/runtime/stack.py", line 137, in run_in_stack
        await run_pulumi_func(lambda: Stack(func))
      File "/home/beer/projects/tacr/platform/cluster-deploy/.venv/lib/python3.11/site-packages/pulumi/runtime/stack.py", line 51, in run_pulumi_func
        await wait_for_rpcs()
      File "/home/beer/projects/tacr/platform/cluster-deploy/.venv/lib/python3.11/site-packages/pulumi/runtime/stack.py", line 121, in wait_for_rpcs
        raise exception
      File "/home/beer/projects/tacr/platform/cluster-deploy/.venv/lib/python3.11/site-packages/pulumi/output.py", line 200, in run
        transformed: Input[U] = func(value)
                                ^^^^^^^^^^^
      File "/home/beer/projects/tacr/platform/cluster-deploy/.venv/lib/python3.11/site-packages/pulumi/output.py", line 250, in lift
        return UNKNOWN if isinstance(v, Unknown) else getattr(v, item)
                                                      ^^^^^^^^^^^^^^^^
    AttributeError: 'NoneType' object has no attribute 'kubeconfig'

pulumi up --skip-preview works fine as expected.

Output of pulumi about

CLI          
Version      3.99.0
Go Version   go1.21.5
Go Compiler  gc

Plugins
NAME           VERSION
gcp            7.4.0
google-native  0.32.0
kubernetes     4.6.1
python         unknown
random         4.15.0

Host     
OS       arch
Version  "23.1.0"
Arch     x86_64

This project is written in python: executable='<REDACTED>/.venv/bin/python3' version='3.11.6'

Backend        
Name         <REDACTED>
URL           <REDACTED>
User           <REDACTED>
Organizations  
Token type     personal

Dependencies:
NAME                  VERSION
pulumi-gcp            7.4.0
pulumi-google-native  0.32.0
pulumi-kubernetes     4.6.1
pulumi-random         4.15.0

Additional context

No response

Contributing

Vote on this issue by adding a 👍 reaction.
To contribute a fix for this issue, leave a comment (and link to your pull request, if you've opened one already).

[mjeffryes]

Thanks for the bug report @jancespivo. Based on the exception, it looks like the container.Cluster constructor isn't returning anything in the failing case.

I notice that you're using a gcp provider that you configure explicitly with the project id, but, although you create a similar google-native provider with the project name, you call containter.Cluster directly without using the google-native provider you configured. I wonder if that's causing the cluster creation to fail?

[jancespivo]

@mjeffryes
There is google-native provider configured:

providers={
     "google-native": google_native_provider,
},

The code I pasted can create Cluster successfully. The only problem is pulumi preview doesn't work.
So pulumi up --skip-preview works, pulumi up doesn't. It fails during preview step.

[mjeffryes]

I apologize, I did miss the provider configuration in the container.Cluster config. The core issue still remains that during preview, the container.Cluster call should return an UNKNOWN, rather than None.