How do I set the VPC for a GCP Cloud Run service? Do the types need to be updated?

Why does Pulumi require the

parameters

for a cloudformation Stack to be a map of string to string (see docs)? Some cloudformation templates use other value types causing them to fail (e.g., "creating CloudFormation stack failed: ValidationError: Parameter 'MinInstancesInService' must be a number."

anyone know how to raise an error condition here?

Hi all, quite new to Pulumi, does anyone know how to provide k8s cluster credentials to Pulumi from a GCP serviceaccount in a

pulumi/pulumi-python

docker container? We are trying to make the following work using Pulumi (highly simplified): 1. Create a k8s cluster on GKE 2. Create a databaseInstance 3. Deploy the database credentials as a secret in the k8s cluster created in step 1 The pulumi/pulumi-python container doesn’t have gcloud-sdk installed (then we could run

gcloud container clusters get-credentials

), this would be an option, but we’d prefer there to be a more ‘native’ way of doing.

I’m trying to use the Pulumi Postgres provider to create databases and roles for an RDS instance provisioned. However I’m getting 

Error initializing PostgreSQL client: error detecting capabilities: error PostgreSQL version: dial tcp :5432: connect: connection refused

 .  But I’m able to connect to the instance using psql. I found some issues via terraform provider, but making similar changes and not having any luck.

I have a list in config that I use to create permissions resources for an azure keyvault I need the permissions to be added before I add keyvault secrets I cannot add the permissions resources to the depends block as they're dynamic Is there a clean way to do this? I could theoretically create a custom resource which creates all the permissions resources and then add that resource to the dependson block but it seems like a lot of work

Q: What is the solution to set a config with a value that starts with

-

? For example:

pulumi config set myValue '---hello'

I would expect the command pass however it yields:

Error: bad flag syntax: ---helo
Usage:
  pulumi config set <key> [value] [flags]

And if I escape the value with

\

, that is `pulumi config set myValue

\-\-\-hello

the set value is not correct Any help is appreciated

Hi, I'm struggling with how to convert an IntOutput to a String for use as an input. Using the Go sdk. Can't seem to find any examples, not sure where to look. In the GCP, I'm capturing the Redis Port from Memorystore which is an IntOutput and want to pass it in a CloudRun template env var, which happens to need to be a String.

Hey, Trying to spin up EKS Getting this error

error: aws:ec2/launchConfiguration:LaunchConfiguration resource 'eks-matan-ng1-nodeLaunchConfiguration' has a problem: Attribute must be a single value, not a list

Hello all, I have pulumi config

config:
  aws:region: us-east-1
  othernamespace:foo bar
  pulumi-config-test:foo: bar

inside my

pulumi-config-test

project. Inside my code:

const pulumi = require('@pulumi/pulumi');

const config = new pulumi.Config();

const foo1 = config.get('foo'); // Outputs bar
const foo2 = config.get('othernamespace:foo'); // undefined
const region = config.get('aws:region'); // undefined

I can't get config variables that are in a different namespace than my project name. So for example, if I want to use

region

variable inside my code I have to set a new config variable:

pulumi-config-test:region

with the same value as

aws:region

variable. Is there any way I can access config variables with a different namespace than the project name inside my code? Thanks for any help!

Hi, all. Looking for some help / insights on

pulumi import

with

SpotFleetRequest

for AWS. I get an error like so:

ec2/spotFleetRequest:SpotFleetRequest resource 'github-runner-spot-fleet-request' has a problem: ExactlyOne: "launch_template_config": one of launch_specification,launch_template_config must be specified

. when importing a spot fleet request definition from AWS. I am importing the launch template as part of the same bulk import.

As part of a stack I deployed an RDSInstance to the wrong account. I’ve since recreated the instance in the proper account. Another stack (via StackReference) now created tables, etc in the proper account. But it’s trying to remove the old tables. Can I export the stack and remove the non-existant tables from the stack? Usually there’s a message suggesting to do so

Help novice looking for some pointers. I am new to Pulumi it is being used to deploy to Azure. We are creating a Windows Appservice and a Linux Appservice, The Linux Appservice has azure functions to ftp to the Windows Appservice, Problem is we would like to store the Credentials for the Windows Appservice in a keyvault that the azure functions can use. Two options appear to be is to somehow access App Credentials that are system generated for the FTPS Endpoint this I believe will not be possible as the AppService might not exist at the time of creating the linux Appservice. Another option would be to create User Credentials when we create the Windows Appservice. Are either options possible in Pulumi and if so where do I need to start looking at. We are utilizing typescript. Thansk

I’m starting to use StackReferences, but unfortunately we followed a pattern where a path element was added to our S3 pulumi state bucket. I’m now going to try to use aws cli tools or similar to remove the path element and merge the state storage together. Not sure if it will work. I’m wondering if anyone has experience with a situation like this. Is there another workaround?

Does

ignoreChanges

support nested properties?

Hi all First of all hello to everyone and thanks for Pulumi which appears to be really great product. I've got question: how to "unimport/detach" resource accidentaly imported to stack? I was looking for it in doc and in this channel but maybe i'm just looking for wrong keyword.

are there any examples for kubernetes on how to create a service, take backup?

Hi all, quick question: I'm trying to use Pulumi to provision a multicloud setup. The AWS and Azure legs are up and working great, but I'm a bit confused by the Pulumi docs for GCP. It might just be that what I want to do isn't available through Pulumi, and I was hoping someone better informed could tell me a quick yes/no whether I can. If so I'll keep trying, and if not I'll need to revisit my setup. I'm trying to provision a GCP Firestore database with Pulumi, but the docs seem to only reference provisioning an index for a db, not the db itself. Am I missing something? I know I can use Google's own SDK for it, but I'd prefer a unified IAC solution

Hi, been using Pulumi on windows, getting this error quite alot could not read plugin "pulumi😛roviders:azure::default_3_41_1': could not read plugin [C:\Users\ChrisThorp\.pulumi\plugins\resource-azure-v3.41.1\pulumi-resource-azure.exe] stdout: EOF" Try uninstalling pulumi and installing via chocolatey but no luck. Anyone got a pointers to fix the issue

For the postgressql provider, I am trying to use

depends_on

to ensure a

Role

exists to perform a

Grant

. But getting:

Exception: 'depends_on' was passed a value that was not a Resource

we keep running into this error any idea what could cause it?

When moving resources to a ComponentResource to improve Code Structure, pulumi wants to delete and create new resources, even though actually only the location in the stack changed, is the same resource with a new parent.

Previewing update (datadog):
     Type                                    Name                                         Plan
     pulumi:pulumi:Stack                     datadog
     ├─ bosch:pt:datadog:DatadogGlobalRoles  DatadogGlobalRoles
 +   │  └─ datadog:index:Role                GlobalReadOnlyRole                           create
 -   └─ datadog:index:Role                   GlobalReadOnlyRole                           delete

Resources:
    + 1 to create
    - 1 to delete
    2 changes. 8 unchanged

This leads to a conflict error:

datadog:index:Role (GlobalReadOnlyRole):
    error: error creating role: 409 Conflict: {"errors": ["A role with the same name already exists"]}

Also the deleteBeforeReplace property is not honoured in this case. Any idea if this is possible without resource recreation?

cross-posted here - not sure if this is the correct channel or not.

What's the recommended way of uipdating Pulumi libs to latest? Currently I visit the Changelogs of each project I'm using and grab the latest versions. Is there a shortcut? A single page or file that lists all the latest versions?

Hello, I'm working with AWS and I'm trying to group my resources (mostly Lambda Functions) in a single Lambda Application, (as described here: https://docs.aws.amazon.com/lambda/latest/dg/deploying-lambda-apps.html ) Is there any way to achieve that using Pulumi?

Hi Everyone, does anyone know, if it is possible to wait for an Azure Function App to become actually available during the pulumi deployment? I need to wait until the function is up and running not only for the availabilty of the resource itself. When creating an Azure EventSubscription, the Azure Function App needs to be up and ready to be able to set an event subscription to the event input binding of an azure function. Does anyone have the same problem?

is there a way in Pulumi to create a SRV record on a ECS Service Discovery service? There seems to be a bug in the documentation for

ServiceDnsConfigDnsRecord

that has

type

values being described as the protocol rather than record type. Also I don’t see any way to add a port for the SRV record. Is this possible?

As anyone else included the use of the AWS-SDK or any other large sdk in a dynamic resource? As soon as I include any function from it I think that pulumi is importing it to the __provider function and the preview just hangs. Been sitting there over 10 minutes for me at this point. This is a test stack so there is barely anything else there. I have debug logs on and its hanging right at the point where it registers resources. If I just comment the function out that step runs in 4 seconds.

My stack is stuck trying to read a stack reference. It worked fine yesterday. From the debug logs it looks like it got the stack reference fine but it just hangs with:

...
I0115 10:15:27.914701   15784 eventsink.go:59] resource read successful: ty=pulumi:pulumi:StackReference, urn=urn:pulumi:prod::tlayen-deployments::pulumi:pulumi:StackReference::/tlayen/test
I0115 10:15:27.914712   15784 eventsink.go:62] eventSink::Debug(<{%reset%}>resource read successful: ty=pulumi:pulumi:StackReference, urn=urn:pulumi:prod::tlayen-deployments::pulumi:pulumi:StackReference::/tlayen/test<{%reset%}>)
I0115 10:15:27.915404   15784 eventsink.go:59] incoming output property translated: name -> name
I0115 10:15:27.915416   15784 eventsink.go:62] eventSink::Debug(<{%reset%}>incoming output property translated: name -> name<{%reset%}>)
I0115 10:15:27.915721   15784 eventsink.go:59] incoming output value translated: /tlayen/test -> /tlayen/test
I0115 10:15:27.915734   15784 eventsink.go:62] eventSink::Debug(<{%reset%}>incoming output value translated: /tlayen/test -> /tlayen/test<{%reset%}>)
I0115 10:15:27.916169   15784 eventsink.go:59] incoming output property translated: outputs -> outputs
I0115 10:15:27.916187   15784 eventsink.go:62] eventSink::Debug(<{%reset%}>incoming output property translated: outputs -> outputs<{%reset%}>)
...
     pulumi:pulumi:Stack      tlayen-deployments-prod  running.    read pulumi:pulumi:StackReference /tlayen/test

Any idea what might be going wrong?