Hey all, I’d like to move some resources between stacks in different pulumi projects. I didn’t see any docs or

pulumi state mv

around, is there any docs on how to do this? Or should I just pulumi state export/import the resources over to the other stack.

Hi there, I have a quick question: is it considered a security risk to push encrypted keys to a public git repository? I am asking because I saw my

Pulumi.dev.yaml

include encrypted secrets after running

pulumi config set ... --secret

(see attached snippet in the thread for details on how the file looks like).

Another short question: I would like to access secrets from my

index.ts

. I tried using

new pulumi.Config();

as explained in https://www.pulumi.com/docs/intro/concepts/config/, but when doing

config.require("name");

it actually prefixes the name with the name of the project

hi, we are currently using a 3rd party rest service and we use it to CRUD secrets. Is creating a dynamic provider the recommended approach in that scenario?

i’m seeing these 2 Route53 errors when creating a certificate similar to the example in

aws-ts-static-website

when running

pulumi up

Error: invocation of aws:route53/getZone:getZone returned an error: transport is closing
Error: invocation of aws:ec2/getSubnetIds:getSubnetIds returned an error: transport is closing

Any advice on how to move past this?

Hello, I'm getting following error on

pulumi preview

and cannot find the cause:

error: Preview failed: unexpected unknown property value

Does anyone have any idea what could be the problem or how I could track down the error?

Question: for CI jobs like github actions, what is the best practice for providing the

secrets.PULUMI_ACCESS_TOKEN

? These seem to be tied to the user, not the pulumi organization.

How do you create a Webhook for Github Enterprise? There is a notion of doing it on the documentation but after reading the source code I don’t see a configuration option to bypass the Github base URL:

Diagnostics:
  github:index:RepositoryWebhook (github-repo-webhook):
    error: POST <https://api.github.com/repos//https://><snip>/security/jupyterhub.git/hooks: 404 Not Found []

Is there an environment variable I can set somewhere?

Here’s my code that resulted in the error above:

github_repo_webhook = github.RepositoryWebhook(
    "github-repo-webhook",
    active=True,
    events=["push"],
    repository='https://<snip>/security/jupyterhub.git',
    configuration={
        "url": codebuild_webhook.payload_url,
        "secret": codebuild_webhook.secret,
        "contentType": "json",
        "insecureSsl": False,
    })

Ah, I believe I found the environment variable here: https://github.com/pulumi/pulumi-github/blob/master/sdk/python/pulumi_github/config/vars.py#L21

Okay new problem, is there anyway to have pulumi-github stop injecting the username for the owner of the supplied github_access_token into the request url?

Diagnostics:
  github:index:RepositoryWebhook (github-repo-webhook):
    error: POST https://<url>/api/v3/repos/<username>/security/jupyterhub.git/hooks: 404 Not Found []

it should just be

https://<url>/api/v3/repos/security/jupyterhub.git/hooks

looks like that would be

owner

env variable… woops

config*

tried setting that and organization to security and got the same thing 😞

In terms of doing updates, are there any situations when cloudformation does it better? For example many RDS updates(such as storage updates) can be done online without downtime with cloudformation. Are there situations where pulumi would require downtime vs cloudformation?

Hello, I have a question on how to reason about Github repositories between stacks, lets say I have a dev and prod environment. A Github repo should only be created once, but I want to use the github repositories to setup build pipelines on both environments.. Is there any handy way to share the github repo resource, but it's only create if it doesn't exist already?

@tall-librarian-49374 Your recent blog post on using container image on AWS lambda with API gateway is great! I would like to know , how I can configure the API gateway in Pulumi to serve a stateless html web application running in a container deployed as a function in AWS Lambda.? i.e. Instead of a json response at the api gateway endpoint, I want to configure the following

"statusCode": 200,

"body":"html",

"headers": {"Content-Type": "text/html"}

. How to use pulumi's AWS API gateway functions that can serve html response from the container image in AWS Lambda ? . It would be great if you could kindly share the approach with the typescript sample code so that we can try it out in our company. Thank you in advance.

Updated to the latest and now I'm getting this error. Appears to maybe related to the fact my repo is on the d: drive. Anyone else have this issue? I killed my credential file all together and reauthenticated, that didn't fix it.

Hello Pulumi team! I was reviewing the Lambda event sources and I noticed that MSK was missing from the list. Is support for this planned?

If I have a repo

Hi all! Sorry, might be a stupid question but I’m struggling to find the answer in the documentation. Tried to check the source code as well for some examples. I am trying to create a custom component in Python. The doc https://www.pulumi.com/docs/intro/concepts/programming-model/#registering-component-outputs tells us to use

self.register_outputs({
    bucketDnsName: bucket.bucketDomainName
})

Which doesn’t work - if it’s a dict, then it should be

"bucketDnsName": bucket.bucketDomainName

and then it should presumably register the output property named “buckerDnsName”? But when I try to use it as

resource.bucketDnsName

it doesn’t work and complains about “AttributeError”. Should I DIY it like

self.bucketDnsName = bucket.bucketDomainName

after self.register_outputs call?

Is there a way to delete a resource from the stack without deleting the cloud resource? Use case: we're importing resources from another account. We can add them to the stack, but we have read-only access in that other account, so we can't delete the resources, nor do we want to. This means that

pulumi destroy

consistently fails, which disrupts the developer workflow. We tried

protect=True

, but that yields another error on

destroy

.

Hi. I am encountering this issue and am wondering if there’s a workaround https://github.com/pulumi/pulumi-awsx/issues/540

Hello, could anyone please post a code snippet, how to subscribe to an Azure EventGrid SystemTopic? I would like to listen on connection-change events from a serverless SignalR Service, but I can't figure out, how to do it with the c# pulumi Azure provider.

I see the page on unit testing pulumi code (https://www.pulumi.com/docs/guides/testing/unit/). But what is the real value add of doing this?

Hi there! Does any know how

pulumi github app

determines whether a status should be reported? I have my

pulumi preview

runs in CI and has every necessary environment variables I thought the integration required (

<http://ci.pr|ci.pr>.number=${github_pr_id}

and

ci.build.type=pull_request

, etc.) , but still cannot get the preview comment and status check update in the Pull Request. Any idea would be highly appreciated🙏

Hi there. I have a problem where

pulumi stack output

and other related commands hangs indefinitely. I've tried all the troubleshooting steps recommended on the website but I've had no joy. Can anyone suggest what I should do next?

Hiya, is there an easy way to get the full urn of a resource in my stack? I have a running Cloud Run service in GCP and just want to force pulumi to replace it with a new image. Having to specify the full urn with --replace seems overly complicated.

In pulumi/aws 3.16.0,

aws.eks.NodeGroup

accepts

instanceTypes

as a

pulumi.Input<string>

. In 3.17.0, a breaking change was made to switch to

pulumi.Input<pulumi.Input<string>[]>

. This results in Pulumi thinking it needs to replace the awseksNodeGroup resource. This is not an acceptable solution. A replacement would result in deleting the node group and then re-creating it, taking the service completely offline. I don't understand how we're supposed to be update to 3.17.0+ with this issue. Are there any solutions to this?