This feature is great

stackConfigDir

but is there any way we can change this variable at runtime? OR have it scan subdirs for configs? (looking to clean up sprawling conf directories)

Hello All, anyone knows how to create pulumi config to gcp, for example, I need create different VM in different regions and zones, how o set different zones and in each zone how to set the number of machines, and how to retrieve that configuration on typescript to execute the creation.

Hi, does anyone know how to check, at runtime, if pulumi is previewing changes vs. executing them? I need to run some code to imperatively create an aws resource using boto3 (a cache invalidation), and it works fine, but when I don't want to create an invalidation during preview mode. So how can I tell in my python program whether it's preview or execute? `I have a project where I want to run some

Hi everyone 👋 We recently had our backend engineer leave our project and we're in a bit of a situation. Our app was removed from Google Play and in order to get it back up we need to deploy via Pulumi. The code is ready to be shipped, but I have not been able to figure out how to deploy to our existing AWS infrastructure and am hoping someone might be able to help out. We're more than happy to pay for someone's time, we really need to get this out ASAP. Apologies if this is the wrong channel or if this isn't allowed! 😅

@full-garage-72020 you can dm me. Will need to know more details on what language, what's the preferred deployment method/tool etc. And how soon?

Hi @everyone I am new to Pulumi and trying to setup my serverless mono app which connects to snowflake pulumi up gives me this error and i am basically stuck here

would appreciate if any of you could provide me with some info

Issues with Azure API Management. Language used c#. I've started with a brand new Azure subscription. There is nothing exotic on that environment. I'm using my personal Azure subscriptions for a PoC. Resource Groups and App insight works. Pulumi Up/destroy all good. The issue is when I add APIM. APIM services and APIM api's don't work. It either takes forever to create APIM or it says it fails but I can see a Resource APIM been created with Status "Activating". Now when I try to do Pulumi destroy, it get's stuck , it can't even delete Resource groups because there is APIM added to it ( even though the previous step has told me that it was unable to create APIM). What I do is Delete stack as well, and then I manually clean up everything. I have done this vicious thing 10 times!! Pulumi Up and Destroy don't seem to be working right when APIM and related services are added. I will try today with something else like CosmosDB.. I'm afraid to touch AKS after this!! I use the scripts I found in https://www.pulumi.com/registry/packages/azure/api-docs/apimanagement/api/ Not sure what's wrong with APIM. Any ideas?

Another problem is the time it takes to provision just 3 resources in Azure using c#: 1. Resource Group 2. App Insights 3. CosmosDB It has created the resources but it took nearly 20 mins.....!! I had deleted the previous stack and created a new one. I'm based in the UK and location was set to uksouth

is pulumi enterprise required to generate an access token that can be used by a CI/CD pipeline like Google Cloud Build?

When I set

parent = Parent(...,ResourceOptions(protect=True))

then pulumi sets

protect=True

on all children as well

child = Resource(..., ResourceOptions(parent=parent)

- is this intentional or do I have any bug in my code?

dns_zone = aws.route53.Zone(
    "zone",
    force_destroy=False,
    name="<http://example.com|example.com>",
    opts=pulumi.ResourceOptions(protect=True),
)

# this is a Component Resource
cert = ACMCertificateR53(  
        cert_config=ACMCertificateR53Config(
            name=f"{PREFIX}-{key}",
            tags=DEFAULT_TAGS,
            zone_id=dns_zone.zone_id,
            domains=["*.<http://example.com|example.com>"],
        ),
        opts=pulumi.ResourceOptions(parent=dns_zone),
    )

I see it being documented here https://www.pulumi.com/docs/intro/concepts/resources/options/parent/ But why is the parent protection passed onto children? I want to protect parent resources only and not child resources.

Could this error be more specific?

Is it possible to prescind of the Pulumi service and use, like in Terraform, say an Azure Blob Storage for persisting the state of the stack? I know this way I should also take care of persisting the secrets generated and all that stuff, but just to know if Pulumi would allow me to be more independent in this way. Also, would like to know if there's any restriction on this aspect, be it technical or monetary.

var nicCoreMediaDevStorageAccount = new StorageAccount("mgcoremedia", new StorageAccountArgs
{
    ResourceGroupName = resourceGroup.Name,
    Kind = "StorageV2",
    
    Sku = new AzureNative.Storage.Inputs.SkuArgs()
    {
        Name = SkuName.Standard_LRS
    },
    Location = "Norway East",
});

Is there a way to set a specific name on a storage account through Pulumi, right now the storage account ends up looking with a generated name appended to the name I have given.

I have a RDS cluster created via Pulumi. I have two postgres db created inside the rds. Now if I want to just destroy one db, Is there any way to do it without destroying the entire cluster .

Hi, could anyone please point me to the procedure of gaining access to your organisation? The current administrator has their account disabled.

Hey guys, any idea when this will be available? https://www.pulumi.com/registry/packages/aws/api-docs/apigateway/restapi/#resources

Hi, I am looking for a way to ignore all changes to a HelmV3 chart once it's been deployed. I have the following code:

const argocd = new k8s.helm.v3.Chart(
      "argocd",
      {
        namespace: "argocd",
        chart: "argo-cd",
        fetchOpts: { repo: "<https://argoproj.github.io/argo-helm>" },
        transformations: [
          (obj: any) => {
            if (obj.apiVersion == "extensions/v1") {
              obj.apiVersion = "<http://networking.k8s.io/v1|networking.k8s.io/v1>";
            }
          },
          args => {
              return {
                  props: args.props,
                  opts: pulumi.mergeOptions(args.opts, { ignoreChanges: ["metadata", "spec"] })
              }
          },
        ],
      },
      { providers: { kubernetes: this.provider }, dependsOn: [argocdPriorityClass]})

This does not ignore changes in the chart child objects - once argocd via pulumi has been deployed, it redeploys another argocd chart with custom values (argocd controlling itself pattern, so i dont want to overwrite this). My terminal reads:

Hi, I would like someone to confirm if such a huge bug can have slipped past Pulumi team before I create an issue to Github. Somehow I find it almost impossible this could have occured but everything I've done is pointing towards that. In Google Cloud when defining IAM bindings those do not get applied even if Pulumi says they do. I've been debugging this for several days now and all IAM bindings I try to apply just don't appear into Google Cloud Console and likewise all operations requiring those fail within our system. I can define those via Google Cloud Console or via

gcloud

as expected and they do appear and everything works. But with Pulumi this simply doesn't happen but in Pulumi state those are set thus leading to Pulumi state not being in sync with the actual environment. The code I'm running has worked without issues before and as such this must be a regression. This happens at least with the latest version of

@pulumi/gcp

(version 6.43.0). I tried downgrading

@pulumi/gcp

to around when our lower environment was created (and when IAM bindings were successfully created) but that didn't help. So it might be something related to newer Pulumi CLI versions or something. I do not know at this point. This however is fatal for us. Not being able to set IAM bindings reliably obviously is more or less catastrophic from IaC point of view. And because of this I'm thinking this kind of bug just cannot have slipped past and I'm reaching out if anyone else has faced this? This worked for us ~2 months ago but not anymore.

hi all! Is it possible somehow to tell Pulumi to ignore changes to a resource only on preview? Context: this resource creation code has some bits that are executed only if it's not

pulumi.isDryRun()

, which means the resource content is different at

preview

and

up

times. I'd like to be able to rely on just the

--expect-no-changes

and not have to parse the diff myself.

The eks package has 3 nodegroups functions (NodeGroup, NodeGroupV2 and ManagedNodeGroup) while the aws classic and native have only 1 (NodeGroup). It seems the ManagedNodeGroup is the one similar to NodeGroup from classic/native. Is that correct? If so, what's the usage to having NodeGroup and NodeGroupV2 on the eks package?

might have run into a small bug in CLI

➜  datadog (wall-port) pulumi version 
v3.47.1
➜  datadog (wall-port) pulumi new --list-templates 
error: could not unmarshal '/Users/wall/.pulumi/templates/kubernetes-azure-yaml/Pulumi.yaml': Configuration key 'azure-native:location' is not namespaced by the project and should not define a type
➜  datadog (wall-port)

Hello everyone, I’ve got a general question about databases. What is your approach to manage things like users and roles? Sure, the script can create the user (login), but then you need to manually assign that user to a role. I am using sqlserver and that has ever been a painful process. Has anyone come up with something that makes things easier to manage?

Hey everyone, im encountering an issue with aws lambda runtime: in the docs i can see that node16js.x is supported:https://www.pulumi.com/registry/packages/aws/api-docs/lambda/function/ yet when im trying to run the code with the same runtime im getting the following error:

aws:lambda/function:Function resource 'campaign_entity_snapshot_function' has a problem: expected runtime to be one of [nodejs nodejs4.3 nodejs6.10 nodejs8.10 nodejs10.x nodejs12.x nodejs14.x java8 java8.al2 java11 python2.7 python3.6 python3.7 python3.8 python3.9 dotnetcore1.0 dotnetcore2.0 dotnetcore2.1 dotnetcore3.1 dotnet6 nodejs4.3-edge go1.x ruby2.5 ruby2.7 provided provided.al2], got nodejs16.x. Examine values at 'Function.Runtime'.

im running pulumi version 3.47.1. What am i doing wrong?

Hey everyone, I’ve got a quick question, is it possible to generate Pulumi schemas from something like an OpenAPI spec? If so, how is this currently done?

Is there a way to customize the resource URNs when writing a dynamic provider? So far I get URNs formatted as:

urn=urn:pulumi:$STACK_NAME::$PROJECT_NAME::pulumi-nodejs:dynamic:Resource::$RESOURCE_NAME

. I'd like to avoid confusion between resources using the same names across different dynamic providers.

hey folks! if anyone is attending re:invent this year, we will have a booth where you can meet some of the team! Please let us know in the thread below!