Hi guys, Sorry if my question is dumb but I'm new to Pulumi 😉 I'm creating a k8s service like:

supertokens_service = k8s.core.v1.Service("supertokens-service",
    metadata={"namespace": namespace.metadata["name"]},
    spec=k8s.core.v1.ServiceSpecArgs(
        selector={"app": "supertokens"},
        ports=[k8s.core.v1.ServicePortArgs(port=3567, target_port=3567)]
    ),
    opts=pulumi.ResourceOptions(provider=k8s_provider)
)

The deployed service has name

supertokens-d8a03ac6

. Two questions: • Can I access the service name using

supertokens_service

variable somehow so I can configure my app container to query this service • Can I set a fixed value for the service to

supertokens

without the suffix so that it remains stable when I do updates, or is it Pulumi behavior to add the suffix and can't be changed ?

Yes to both:

supertokens_service = k8s.core.v1.Service("supertokens-service",
    metadata={"namespace": namespace.metadata.name,
              # set a fixed name instead of auto-naming
              "name": "supertokens"},
    spec=k8s.core.v1.ServiceSpecArgs(
        selector={"app": "supertokens"},
        ports=[k8s.core.v1.ServicePortArgs(port=3567, target_port=3567)]
    ),
    opts=pulumi.ResourceOptions(provider=k8s_provider)
)

# reference the service's name in another resource
other_resource = Resource("other-resource",          service_name=supertokens_service.metadata.name)

# export the service's name as a stack output for downstream use
pulumi.export("supertokens-service-name", supertokens_service.metadata.name)

Ty ty 😉 So I tried setting a variable like:

supertokens_service_name = supertokens_service.metadata['name']

For my other resource, it's:

auth_api_deployment = k8s.apps.v1.Deployment("auth-api",
    metadata={"namespace": namespace.metadata["name"]},
    spec=k8s.apps.v1.DeploymentSpecArgs(
        selector=k8s.meta.v1.LabelSelectorArgs(
            match_labels={"app": "auth-api"}
        ),
        replicas=1,
        template=k8s.core.v1.PodTemplateSpecArgs(
            metadata={"labels": {"app": "auth-api"}},
            spec=k8s.core.v1.PodSpecArgs(
                containers=[k8s.core.v1.ContainerArgs(
                    name="auth-api",
                    image=auth_api_image,
                    env=[
                        k8s.core.v1.EnvVarArgs(
                            name="SUPERTOKENS_URL",
                            value=f"http://{supertokens_service_name}:3567"
                        ),
                        k8s.core.v1.EnvVarArgs(
                            name="SUPERTOKENS_API_KEY",
                            value=supertokens_api_key
                        )
                    ]
                )]
            )
        )
    ),
    opts=pulumi.ResourceOptions(provider=k8s_provider)
)

However it seems I can not use

supertokens_service_name

like this, as a value i'm getting:

<http://Calling> __str__ on an Output[T] is not supported.

            To get the value of an Output[T] as an Output[str] consider:
            1. o.apply(lambda v: f"prefix{v}suffix")

            See <https://www.pulumi.com/docs/concepts/inputs-outputs> for more details.
            This function may throw in a future version of Pulumi.:3567
        - name: SUPERTOKENS_API_KEY

Welcome to Pulumi Outputs 😉

Haha, should I do:

supertokens_service_name = supertokens_service.metadata['name'].apply(lambda name: str(name))

?

Yes, this is one option. You need to keep in mind that the value of the output is not known at the point in time the Python code is evaluated. So you need to keep it as a Python object at all times, you cannot convert it to a string at runtime.

Ah I think I get it, so this should work ?

env=[
                        k8s.core.v1.EnvVarArgs(
                            name="SUPERTOKENS_URL",
                            value=pulumi.Output.format("http://{0}:3567", supertokens_service_name)
                        ),

Yes, that should work

perfect !

As a small service I'll also tell you that

pulumi.Output.json_dumps

is a thing. Not relevant here, but something that I would have liked to discover earlier. It often saves you from insane

.apply()

constructs.

Nice ! That's what I hated about Terraform tbh, insane one-liners to extract something from a data struct

> Love Pulumi so far, once I get around these hurdles I will drop Terraform completely I only started using Pulumi last year, and I love it as well, hands down the best IaC framework I've used so far. I found that once I understood how outputs work (generally and in Python) it was very easy and convenient to use

That's great ! Pulumi GPT seems to not be aware of it =D

No, and be careful with this tool, it will happily invent APIs that look absolutely plausible and seem to be just made for your particular purpose, but in reality don't exist 😄