I’m having problems upgrading from pulumi 2 to 3 (and upgrading the plugin dependencies). After upgrading the CLI and changing the versions in

package.json

and removing

node_modules

and calling

yarn install

again, when I run

pulumi preview

, I get:

aws:iam:Role (eks-cluster-node-group-role):
    error: could not validate provider configuration: 1 error occurred:
    	* Invalid or unknown key

followed by a bunch of stack traces that look like this:

Error:
        at Object.debuggablePromise (/scratch/node_modules/@pulumi/pulumi/runtime/debuggable.js:69:75)
        at /scratch/node_modules/@pulumi/pulumi/runtime/invoke.js:126:45
        at Generator.next (<anonymous>)
        at fulfilled (/scratch/node_modules/@pulumi/pulumi/runtime/invoke.js:18:58)
        at processTicksAndRejections (node:internal/process/task_queues:96:5)
    unhandled rejection: CONTEXT(6017): Invoking function: tok=kubernetes:yaml:decode asynchronously
    STACK_TRACE:
    Error:
        at Object.debuggablePromise (/scratch/node_modules/@pulumi/pulumi/runtime/debuggable.js:69:75)
        at /scratch/node_modules/@pulumi/pulumi/runtime/invoke.js:126:45
        at Generator.next (<anonymous>)
        at fulfilled (/scratch/node_modules/@pulumi/pulumi/runtime/invoke.js:18:58)
        at processTicksAndRejections (node:internal/process/task_queues:96:5)

unfortunately this error isn’t very useful Any hints?

Hi All, I'm interested in giving pulumi a try. To get started I'm trying to get tf2pulumi going but it seems to ignore my *.auto.tfvars files. Am I missing an option to get that working or is there a simple way to get my variables imported?

Hello, I have been facing some odd is issues with Pulumi and GCP Uptime Check we added recently uptime checks to our pulumi stack, we create the uptime check like this

ucc, err := monitoring.NewUptimeCheckConfig(ctx, name, &monitoring.UptimeCheckConfigArgs{
		DisplayName: pulumi.String("uptime check example"),
		HttpCheck: &monitoring.UptimeCheckConfigHttpCheckArgs{
			Path:          pulumi.String(fmt.Sprintf("/%s/status", "github")),
			Port:          <http://pulumi.Int|pulumi.Int>(443),
			RequestMethod: pulumi.String("GET"),
			UseSsl:        pulumi.Bool(true),
			ValidateSsl:   pulumi.Bool(true),
		},
		MonitoredResource: &monitoring.UptimeCheckConfigMonitoredResourceArgs{
			Labels: pulumi.StringMap{
				"host": pulumi.String(targetUrl),
			},
			Type: pulumi.String("uptime_url"),
		},
		Period:  pulumi.String("60s"),
		Timeout: pulumi.String("10s"),
	})

This was working fine, but then we decided to add an alert policy for this uptime check Note: here we forward the uptime check created previously

args := monitoring.AlertPolicyArgs{
		DisplayName: pulumi.String(name),
		Combiner:    pulumi.String("AND"),
		Conditions: monitoring.AlertPolicyConditionArray{
			monitoring.AlertPolicyConditionArgs{
				DisplayName: pulumi.String("Health check alerts for github %s", service.ShortName),
				ConditionThreshold: monitoring.AlertPolicyConditionConditionThresholdArgs{
					Filter:   pulumi.Sprintf("metric.type=\"<http://monitoring.googleapis.com/uptime_check/check_passed\|monitoring.googleapis.com/uptime_check/check_passed\>" AND metric.label.check_id=\"%s\" AND resource.type=\"uptime_url\"", uptimeCheck.UptimeCheckId),
					Duration: pulumi.String("60s"),
					Trigger: monitoring.AlertPolicyConditionConditionThresholdTriggerArgs{
						Count: pulumi.IntPtr(1),
					},
					ThresholdValue: pulumi.Float64Ptr(1),
					Comparison:     pulumi.String("COMPARISON_LT"),
					Aggregations: monitoring.AlertPolicyConditionConditionThresholdAggregationArray{
						monitoring.AlertPolicyConditionConditionThresholdAggregationArgs{
							AlignmentPeriod:  pulumi.String("60s"),
							PerSeriesAligner: pulumi.String("ALIGN_COUNT_TRUE"),
						},
					},
				},
			},
		}
		NotificationChannels: "alerts",

This worked fine in the first deployment, but the subsequent ones started to fail.

error: deleting urn:pulumi:env::company::gcp:monitoring/uptimeCheckConfig:UptimeCheckConfig::uptime-check-github: 1 error occurred:
Error when reading or editing UptimeCheckConfig: googleapi: Error 400: Request contains an invalid argument.

What a noticed is the new uptime checks got created in our account, but GCP entered in some weird state where it could not delete the previous uptime check, the only way I managed to fix the stack was by deleting the old uptime checks manually, I also noticed that the old uptime checks still had alert policies attached to them Is there any idea of what could be happening? I personally can’t find what I have missed and there’s not much documenation around pulumi and uptime checks.

I was wondering if somebody could help me out with creating some Azure WebApps using Pulumi… I want them to run dotnet, and tried that with the following code:

const webapp = new web.WebApp(
		projectKey,
		{
			reserved: true,
			resourceGroupName: resourceGroup.name,
			serverFarmId: appServicePlan.id,
			siteConfig: {
				linuxFxVersion: "DOTNET|6.0",
				netFrameworkVersion: "v6.0",
			},
		}
	);

	new web.WebAppMetadata("meta", {
		resourceGroupName: resourceGroup.name,
		name: webapp.name,
		properties: {
			CURRENT_STACK: "dotnetcore",
		},
	});

The

linuxFxVersion

kind of worked, which means that I can create a deployment in Azure’s Deployment Center that seems to recognize dotnet as runtime. However, if I then deploy a .net app, it seems to me as if the application is started like a PHP application, which obviously does not work… All I see is an apache forbidden page. I tried to add the

WebAppMetadata

with the

CURRENT_STACK

and the

netFrameworkVersion

, but that didn’t help neither, so I am wondering if that is necessary at all… Another thing is that the general settings tab within configuration in portal.azure.com has no stack selected, although I tried to configure it using the code above. Can anybody tell me what I am doing wrong?

is there a way to skip a resource with

pulumi up

? sort of the inverse of

--target

Hey. I am mounting filestore in gce, (with the correct firewalls) the GCE and filestore are created through pulumi but the mounting is giving a timeout error. When I am creating through the console, the mounting works. Any solution?

Hi, I've noticed that the https://github.com/pulumi/examples repository has the Apache-2.0 License, however some files have the following

Copyright

header.

// Copyright 2016-2019, Pulumi Corporation.  All rights reserved.

Am I allowed to use the examples under Apache-2.0 terms and conditions?

The following code seems kind of ugly:

var searchDns = new PrivateZone("<http://privatelink.search.windows.net|privatelink.search.windows.net>", new PrivateZoneArgs
        {
            PrivateZoneName = "<http://privatelink.search.windows.net|privatelink.search.windows.net>",
            ResourceGroupName = (
                _coreRef.GetOutput("CoreResourceGroupName") 
                    ?? throw new Exception("Core resource group not found"))
                .Apply<string>(rg => Convert.ToString(rg) 
                    ?? string.Empty)
        });

Am I doing something wrong?

_coreRef.GetOutput()

always returns

Output<object?>

but ResourceGroup name is

Input<string>

. Is there an easier way to convert to

Input<string>

from the output of

GetOutput()

?

hi, for rancher2 provider, i wonder if anyone has successfully used

custom_cloud_provider

to provision a cluster on cloud providers (other than aws/gcloud/azure etc. that are already supported)? https://www.pulumi.com/registry/packages/rancher2/

another question about rancher2 provider, since this provider requires you have a rancher k8s cluster in place. this brings a chicken-and-egg question, how do you create the cluster for rancher (following IasS best practices) in the first place?

Hi, we are having this issue with eks deploy https://github.com/pulumi/pulumi-eks/issues/676 Any ideas how to resolve?

Hi, I am trying to deploy a VM on VMware vSphere and then run a shell script from local (not remote). How do I run the shell script where I need to pass multiple arguments to it? This is what I am trying it doesn’t seem to work

script_output = command.local.Command("execute_shell_script",
                       create='chmod 755 configure.sh && ./configure.sh \"{pulumivm.default_ip_address}\" {arg1} {arg2}',
                       opts=pulumi.ResourceOptions(depends_on=[pulumivm])
                       )
pulumi.export("script_output", script_output.stdout)

Can you please point me to a snippet or an example where this is being done?

👋 Hi everyone!

Hello, is there a way to get the information during runtime if the resource will be updated? I am using python

hi community👋, i know a lot of folks use pulumi (infra provisioning) as well as flux (for gitops). i am curious why there isn't a pulumi provider for flux given there is a terraform provider. i imagine we can use flux pulumi provider to deploy flux onto the cluster.

Hello Everyone! My name is Kelvin DeCosta. I'm very new to using IaC tools and so far my experience has been great! My team and I have a few tasks that we would like to automate. We need to execute some long-running scripts at the start of each week. We've settled on using

AWS Batch

with

Fargate

. We've tried and tested a "dispatcher"

Lambda

that is invoked at the start of each week (via

EventBridge

/

CloudWatch

events, in

TypeScript

). This

Lambda

would submit

Jobs

with a specified

Job Definition

to a

Job Queue

which would then schedule them accordingly. I've set the

Lambda

function to use a

Docker

container the image of which is hosted on

ECR

(

Docker

container runs a

Python

script). I'm having a little trouble with connecting the

AWS Batch

resources (

ComputeEnvironment

,

JobDefinition

, etc.) and

ECS

,

Fargate

. I would really appreciate it if I can get some help/advice on linking these resources together. AFAIK, there are no examples using

AWS Batch

and

Fargate

with

Docker

containers in the examples repository. Feel free to correct me where I'm wrong. Any help is appreciated. Thank you for your time!

Hey, I'm using the MongoDB Atlas package with Pulumi and I'm getting this error:

debug: setting state: Invalid address to set: []string{"snapshot_backup_policy", "0", "policies", "0", "policy_item", "0", "id"}
error: update failed

It seems to be successfully setting everything up and then it's gets the new cluster and appears to try to set it to the pulumi stack state and fails. Is this something I'm doing incorrectly?

Hey guys. I am new around here and on Pulumi. Coming from a Terraform provider development background. I am curious to know what materials, videos, etc you recommend as a starting point, to start developing a Pulumi provider. Thanks

Hi folks, looking for a bit of help with the login capability. Is there a way to run the pulumi login from the automation API ?

do pulumi stack needs to connect to specific aws profile to work with respective aws environment aka dev/integration/prod.... how do pulumi stack know which env to work?

I dont see anything related to canaries in cloudwatch, can you please help me.

any way to get pulumi to re-output the generated code from a

pulumi import

? i imported a lot of things using TMUX and...yeah

Hi All, just quick check. Is there anyway to hard code an stack with azure subscription id?

Hey everyone 👋 I’m using F# in a Pulumi project and tried importing a GCP Composer environment. Since the code defaulted to C#, I had to translate it to F#. It’s now in a state where it builds fine but when I run

pulumi preview

it complains about something that looks like missing properties - but I’m not sure why, as those properties were not included in the C# code that was generated by the import command. Any tips for what to look for?

I would be very happy if the people of this community can tell about themselves too. First of all, I want to tell about myself, my name is Bungla, and I am the director of adsansar.com.

Hi guys, I am relatively new on IAC. I would wish to ask for opinions on the best practice to organize Pulumi projects. This link provides the best reference so far. However, I do not like the way that the app's main.go goes into the same repository with the other Pulumi code as it could reduce visibility. I wonder it would be possible to place the app's main.go/index.ts together with the application code. I have an example in the picture attached. This allows the pkg to be referenced from the service_api repo while having the visibility required. Please feel free to comment on this structure.

Hi, consider the following

Pulumi.<stack>.yaml

file:

config:
  config:one:
    key: val
  config:two:
    key: val

In order to read this config I have to:

config = pulumi.Config(name="config")
config_one = config.require_object("one")
config_two = config.require_object("two")

Is there away to

for

loop`config`over each object? Thanks

Hi all, I'm new in the Pulumi world, so forgive me if it is a stupid question. Our current VPC along is set up through

awsx.ec2.Vpc

The issue I can't figure out is if there is a way to edit the route table that it has created in the background. I found documentation on importing and creating new route tables, but nothing on how to edit the existing route tables and if that is possible when you create the VPC using the Pulumi Crosswalk.

Hey folks, Is there a simple way to expect specific changes and fail if there are more. This is sort of similar to expect no change. For example I want to say only expect image version change, anything more it should not continue.