Hi everyone! Question regarding azure, how can I manually create an onEvent to a trigger through an

ArchiveFunctionApp

?

👋 Hi everyone!

I'm working on setting up a new AWS build pipeline as exampled here: https://www.pulumi.com/docs/guides/continuous-delivery/aws-code-services/#buildspecyml Can someone point me the right direction, I'm getting an error on the BUILD phase that is kind of vague. "COMMAND_EXECUTION_ERROR: Error while executing command: update_pulumi_stack.sh. Reason: exit status 127"

How are people structuring there pulumi Go project folders? What directories do you use? Go recommends folders like CMD, Internal.

Here is the build log for where it failed at: [Container] 2022/03/15 15:56:47 Entering phase BUILD [Container] 2022/03/15 15:56:47 Running command npm install npm WARN read-shrinkwrap This version of npm is compatible with lockfileVersion@1, but package-lock.json was generated for lockfileVersion@2. I'll try to do my best with it!

node scripts/install-pulumi-plugin.js resource aws v4.38.1

[resource plugin aws-4.38.1] installing Downloading plugin: 0 B / 96.61 MiB 0.00% Downloading plugin: 6.44 MiB / 96.61 MiB 6.66% 2s Downloading plugin: 14.27 MiB / 96.61 MiB 14.77% 2s Downloading plugin: 21.61 MiB / 96.61 MiB 22.37% 2s Downloading plugin: 28.77 MiB / 96.61 MiB 29.77% 1s Downloading plugin: 35.59 MiB / 96.61 MiB 36.83% 1s Downloading plugin: 42.65 MiB / 96.61 MiB 44.14% 1s Downloading plugin: 49.28 MiB / 96.61 MiB 51.01% 1s Downloading plugin: 61.25 MiB / 96.61 MiB 63.40% Downloading plugin: 77.83 MiB / 96.61 MiB 80.55% Downloading plugin: 92.16 MiB / 96.61 MiB 95.39% Downloading plugin: 96.61 MiB / 96.61 MiB 100.00% 2s

node scripts/install-pulumi-plugin.js resource kubernetes v3.16.0

[resource plugin kubernetes-3.16.0] installing Downloading plugin: 0 B / 30.27 MiB 0.00% Downloading plugin: 7.33 MiB / 30.27 MiB 24.22% Downloading plugin: 14.56 MiB / 30.27 MiB 48.09% Downloading plugin: 30.27 MiB / 30.27 MiB 100.00% 0s

node scripts/postinstall

added 150 packages from 235 contributors and audited 153 packages in 8.304s 33 packages are looking for funding run

npm fund

for details found 0 vulnerabilities [Container] 2022/03/15 15:57:00 Running command npm run build npm ERR! missing script: build npm ERR! A complete log of this run can be found in: npm ERR! /root/.npm/_logs/2022-03-15T15_57_00_514Z-debug.log [Container] 2022/03/15 15:57:00 Command did not exit successfully npm

Issue resolved. There is a service role created for the specific build project. You must add the appropriate IAM permissions to that account to CRUD your infrastructure. Also buildspec.yml was cleaned up and simplified to be just this:

version: 0.2

phases:
  install:
    commands:
      # pulumi
      - curl -fsSL <https://get.pulumi.com/> | sh
      - export PATH=$PATH:$HOME/.pulumi/bin
  build:
    commands:
      - npm install
      - pulumi stack select dev
      - pulumi up --yes

Hello everyone, I hope everything is OK, I have this error using dynamic provider with

true-myth

library

pulumi:pulumi:Stack (wordpress-dev):
    TypeError: Symbol.prototype [ @@toPrimitive ] requires that 'this' be a Symbol: TypeError: Symbol.prototype [ @@toPrimitive ] requires that 'this' be a Symbol
        at Symbol.[Symbol.toPrimitive] (<anonymous>)
        at Object.<anonymous> (<anonymous>:64:21)
        at Module._compile (node:internal/modules/cjs/loader:1099:14)
        at requireFromString (/home/abukamel/codebases/work/iac/pulumi/programs/wordpress/node_modules/require-from-string/index.js:28:4)
        at getProvider (/home/abukamel/codebases/work/iac/pulumi/programs/wordpress/node_modules/@pulumi/cmd/dynamic-provider/index.ts:66:20)
        at Object.<anonymous> (/home/abukamel/codebases/work/iac/pulumi/programs/wordpress/node_modules/@pulumi/cmd/dynamic-provider/index.ts:116:26)
        at Generator.next (<anonymous>)
        at /home/abukamel/codebases/work/iac/pulumi/programs/wordpress/node_modules/@pulumi/pulumi/cmd/dynamic-provider/index.js:21:71
        at new Promise (<anonymous>)
        at __awaiter (/home/abukamel/codebases/work/iac/pulumi/programs/wordpress/node_modules/@pulumi/pulumi/cmd/dynamic-provider/index.js:17:12)
 
    error: Symbol.prototype [ @@toPrimitive ] requires that 'this' be a Symbol

The code works normally if it was not used directly without pulumi.

export const getOrCreateMonitor = async (fqdn: string): Promise<Result<Monitor, string>> => {
    await checkApiKey()
    const currentMonitor = await getMonitor(fqdn)
    if (currentMonitor.isOk) {
        return currentMonitor
    }
    const alertContacts = await getAlertContacts()
    if (alertContacts.isErr) {
        return err(alertContacts.error)
    }
    const formattedAlertContacts = alertContacts.value.map((alertContact: AlertContact) => `${alertContact.id}_5_0`)
    const requestData =
        qs.stringify({
            ...sharedRequestData,
            type: '1',
            alert_contacts: formattedAlertContacts.join('-'),
            friendly_name: fqdn,
            url: `https://${fqdn}`
        })
    const response = await <http://axios.post|axios.post>(newMonitorUrl, requestData, sharedRequestConfig)
    if (response.data.stat != 'ok') {
        return err(response.data.error.message)
    }
    const newMonitor: Monitor = response.data.monitor
    return ok(newMonitor)
}

Is there a way to fix this?

I had a question on this: https://github.com/pulumi/pulumi/issues/8796 I was wondering if for this (and possibly other breaking changes), if pulumi could implement feature flags so that users could “subscribe” to these types of bug fixes?

Hi, I have been using Pulumi in the past 2-3 weeks to setup our infrastructure on AWS and it’s been great. We are considering setting up Amazon Managed Grafana. Wondering if there are any plans to configure Amazon Managed Grafana workspace using pulumi in the near future (apologies if I missed any discussion regarding this).

Hi everyone,

How do you create a mock resource? (in golang)

I just want some arbitrary resource that fits the

pulumi.Resource

interface

Just wondering if anyone else got this error, which only happens when I run pulumi up (refresh and preview are successful) from a CI/CD pipeline: Error: failed to register new resource …… [aws:cloudfront/distribution:Distribution]: Resource monitor is terminating The actual resource is not always the same, or the same type. Running the same commands from my local, but the same code branch, is successful.

Hi all! Could anyone help me with a pricing question please? I am responsible for a very small team and we don't have a lot of free funds at this stage. Looking through the pricing page on the Pulumi site, it seems as though, if we are self-hosting the backend/state file, there isn't much need to pay for a teams licence. Is there something that I am missing here?

What would be the most pain-free way of migrating from a monolithic stack to smaller service-based stacks? Ie. Is there an easy way to migrate state if I am already making use of a parent relationship for the "services" which I want to pull out of a monolithic deployment?

Anyone have any examples of using aws.ssm.Association with an

Automation

document (not Command)? Also is there a way of 'triggering' Automation documents without associating to instance (without dropping to command line) for example to setup a DB or Active Directory users.

Does anyone have an example of GKE private cluster?

Hi all, I see Google Cloud Native is recommended for new projects. Does it support more resources vs the old Google Cloud Classic?

My Pulumi stack is in a broken state that I believe should be impossible, not sure what I’m missing. I have a stack with multiple AWS providers for different AWS accounts. I deactivated one of these accounts outside of Pulumi which, as expected, caused

pulumi up

and

pulumi refresh

to stop working. So I exported the stack, edited the JSON to remove all references to resources or providers in the removed account, and imported the stack. Now,

pulumi refresh

works fine, and the stack looks right in the UI and when I export it and look at the JSON. But when I try to run

pulumi up

it still complains because it tries to assume a role in the old account to refresh a resource that no longer exists in the stack. The resource referred to in the error message is not present in the UI, using

pulumi stack --show-urns

or when exporting and grepping the

stack.json.

How can this be? Why is Pulumi still trying to access that resource that it shouldn’t even know exists at this point?

I am trying to create a GKE cluster with the network I created but I keep getting this error even though I have created NAT gateway for the network:

Network "network" has no route to the default internet gateway. To fix this, add a route with destRange '0.0.0.0/0' and nextHopGateway 'default-internet-gateway'.

I was able to create a GKE cluster using this network via google console. Does anyone know what the fix is? Thank you

Hey Output.toString in nodejs sdk still silently fails and returns the error message, for the past 3 years it has said "This function may throw in a future version of @pulumi/pulumi.`;". Is there any reason this hasn't been made to throw, surely this causes countless more headaches than just throwing it. I've seen PRs introducing this same weird behaviour in Python and C# sdks such that its become a pattern.

Hello! Am I understanding correctly that self-managed backends don't namespace stacks by project? https://github.com/pulumi/pulumi/issues/2522

Hi, AFAIK Google Cloud Native is in pre-release state now. Do you know when it's gonna be ready for production use?

Hello. Does anyone know how to implement the suggested approach from the error message below in golang ?

./main.go:123:3: cannot use arn (type pulumi.StringOutput) as type pulumi.StringMapInput in field value:
    	pulumi.StringOutput does not implement pulumi.StringMapInput (missing ToStringMapOutput method)

    error: an unhandled error occurred: program exited with non-zero exit code: 2

I am trying to create a Kubernetes secret from a secret stored in AWS Secrets Manager. I can lookup the secret in SecretsManager but I don;t know how I can pass the secret data in to the kubernetes secret pulumi module.

func CreateK8sSecret(ctx *pulumi.Context, arn pulumi.StringOutput, label pulumi.StringMap, provider *kubernetes.Provider) error {
	found_secret := secretsmanager.LookupSecretOutput(ctx, secretsmanager.LookupSecretOutputArgs{
		Arn: arn, # Don't know how I could pass the secret data in
	}, nil)

	ctx.Export("kubernetes_secret", found_secret)

	corev1.NewSecret(ctx, "pulumi-secret", &corev1.SecretArgs{
		// Data: pulumi.StringMap{"pulumi": pulumi.String("new environment")},
		Data: arn,
		Metadata: &metav1.ObjectMetaArgs{
			Labels: label,
		},
	}, pulumi.Provider(provider))

	return nil
}

Hey all, I was wondering if someone else got this behaviour. Running pulumi up, the preview shows something different than what the update will actually do. For example the preview: 2022-03-16T07:24:25.3394059Z ~ 7 to update 2022-03-16T07:24:25.3394311Z - 1 to delete 2022-03-16T07:24:25.3394599Z 8 changes. 376 unchanged But the update finishes successfully reporting this: 2022-03-16T07:27:10.8394827Z Resources: 2022-03-16T07:27:10.8395196Z ~ 2 updated 2022-03-16T07:27:10.8395579Z 382 unchanged 2022-03-16T07:27:10.8395722Z 2022-03-16T07:27:10.8395984Z Duration: 2m8s Is this normal?

I'm getting an internal assertion thrown here: https://github.com/pulumi/pulumi/blob/8b516bb05434ca8c6dac94b76a5583e855ffd0b6/sdk/go/common/resource/plugin/provider_plugin.go#L808 (Actually, it's line 807 in the stack trace, but presumably either the URN or ID is empty..) This is in a brand new project/stack. Never been deployed. Any ideas on what I should be looking for?

By chance has anyone setup AWS with Okta SSO, and are managing group/policy assignments with Pulumi? In order to assign a Okta Push Group to an AWS account, one has to know the Group ID that AWS created when the group was pushed from Okta. I don't see a way to get that ID with Pulumi and really don't want to have to hard code each group.

Hi everybody, Is there a way to change a resource ID without hacking the stack state or bastardizing resources with aliases? If I could manage this using the CLI, that would be great. Many thanks!

Hey all, has anyone had any chance using

stageVariables

to define AWS API Gateway integrations? I have been trying to do something like this:

new Aws.ApiGatewayV2.Integration("lambda", new Aws.ApiGatewayV2.IntegrationArgs
{
    ApiId = "...",
    IntegrationMethod = "POST",
    IntegrationType = "AWS_PROXY",
    IntegrationUri = "${stageVariables.apiFunc}",
    PayloadFormatVersion = "2.0",
});

but pulumi is complaining the

IntegrationUri

is not a proper ARN