you can

tsc

it fine and it will use the link, but when pulumi rebuilds the typescript it ignores the links

and it looks straight at the 'real' module instead

it is something i'd love to get working, it would be a great help

Interesting - we use

yarn link

pretty heavily in our development (or at least, I do!)

i guess i'll try yarn - never used it

Could you open an issue with repro steps for

npm link

if you have them though? I think that’s something we’ll want to fix up

(Our templates default to

npm

so it should probably work well)

yeah will do, wasn't sure if it was meant to be supported

Contrib docs mention yarn explicitly, so I got to learn yarn link and it works quite good

When writing in Pulumi JS, do you have to write in the order that they depend on each other? Or does Pulumi calculate a dependency graph behind the scenes regardless of the order you write your infrastructure components in?

You will mostly end up writing things in order because of scoping though

Hello, I got a question for setting desired node for node groups in EKS. I'm trying to get Cluster Autoscaler working in our EKS cluster, which means the actual desired nodes may vary from the configuration we set in Pulumi, which result in doing an pulumi update next time will set the node count back to what we've configured. Is there a way to avoid this? Cuz I also noticed Pulumi will set desired node to 2 if we didn't pass in the desired count.

Hi guys I am trying the demo to deploy a site using s3, it works like a charm but now sure what are the required adjustments when I want to deploy a more realistic site, for example a folder with other directorios inside like images, js, css, etc. I am getting the following error in this case:

Error: failed to register new resource media [aws:s3/bucketObject:BucketObject]: 2 UNKNOWN: failed to compute asset hash: asset path 'www/static/media' is a directory; try using an archive
        at Object.registerResource

Can you deploy the k8s dashboard to minikube using Pulumi?

Hi there, following some 403 errors which at the end seemed related to my vpn while doing some

helm fetch

, I just noticed pulumi actually do these

helm fetch

for every chart at each run while there’s no need once they are installed and the version did not bump. Is this not overkill and if only triggered on a version change could reduce the time a

pulumi up

takes ?

Hi there, I'm new at pulumi. At the momentan I'm using pulumi with typescript on azure. I want to do create a configuration file, which include all my parameter values. What I did is, I created stack file for a test environment with some keys and values. All of them are strings. If I start pulumi with

pulumi config --config-file Pulumi.Test.yaml

to check my values, but the no keys or values are visible. My yaml file looks like: `encryptionsalt: ****** template: config: azure:environment: description: The Azure environment to use (

public

,

usgovernment

,

german

,

china

) default: public project:name: default: TestProject project:location: default: West Europe`

I'm having an issue in CI with

error: could not load plugin for aws provider

. I have updated the AWS provider version in my project, but the state file still has an older version referenced which does not get installed. CLI is version

0.17.11

. It seems there were updates to try and prevent this, installing any providers on

pulumi up

, but it seems to be failing here. I am using the S3 remote state capability, I wonder if it could be a bug with this?

Hello there, I don’t really know where the problem comes from, and I’m pretty sure that’s not pulumi related, but I could use some insights. I’m unable to pull images anymore from GCR on my GKE cluster every 45 days or something without any reason or changes on its configuration or service accounts. The GCR bucket is in the same project as the GKE cluster and everything works fine up until around 45 days. So far I had to recreate the cluster from my scripts to get rid of the problem, but that’s overkill. I suppose some auth keys expires or something but I can’t really figure it out. Does anyone know what I should do or where to look at ? thanks !

@tall-librarian-49374 by any chance are you around atm?

Hello from Colombia, i am using pulumi(beginner level) to deploy resources in AWS , yesterday i had a meetup where i presented some features of Pulumi, a lot of people was interested with this tool but their principal question was : what happen with state of resource if pulumi web fail? i understand that pulumi store the state of resource into the web.

Pulumi have some SLA where define that ?

seems like

pulumi logs

returns no logs for functions that were created with an aws provider that does assume-role - is this a known issue? is there a workaround?

Does pulumi have something similar to

Packer

?

I'm creating a

pulumi.ComponentResource

which internally creates some

aws.cloudformation.Stack

. I would like to expose some outputs of the CF Stacks as outputs of my

ComponentResource

. I tried doing this in my ComponentResource ctor:

this.computeGroupName = computeStack.outputs.apply(outputs => outputs["CodeDeployDeploymentGroup"]);

Typescript complains that

Type 'Output<any>' is not assignable to type 'Output<string>'.

. I'm guessing this is because a CF Stack does not cast the values from its

.outputs

to any type.

this.computeGroupName

is defined as

computeGroupName: pulumi.Output<string>;

. Is there a way to make TypeScript happy?

Can you nest a

pulumi.ComponentResource

within another

pulumi.ComponentResource

? If so, should all ComponentResouces take a

pulumi.ComponentResourceOptions

map?

Did something recently changed with regards to precedence rules of provider attributes in the kubernetes (typescript) provider? I recently upgraded to

@pulumi/kubernetes: 0.23.1

and seeing some odd behaviour: We instantiate provider instances dynamically via

new gcp.Provider({kubeconfig: ..., namespace: 'namespace-A'})

and then use this provider instance in some resources like

new k8s.core.v1.ServiceAccount(foo, {metadata: {namespace: 'namespace-B'}}, {provider: k8sProvider})

. I would expect the serviceaccount to be created in

namespace-B

in the example above, but it actually gets created in

namespace-A

. I can create a more detailed repro if necessary but wanted to ask first if this intentional or a known issue? @creamy-potato-29402?

Hey i'm getting an error with this code:

This is the error im getting:

error: aws:lambda/function:Function resource 'pulumi-test-lambda' has a problem: "vpc_config.0.vpc_id": this field cannot be set

vpc

is a vpc object. Am I not able to throw in the vpc.id object into the configs?