anyone have a good example of AWS ECS connected to RDS using IAM auth?

I've the following code

const asgEventsTopic = new aws.sns.Topic("asg-events-topic")

asgEventsTopic.onEvent("ci-cd-scale-in", async ev =>{
    console.log("Processing " + ev);
})

const runnerAsg = new aws.autoscaling.Group("ci-cd-runner-asg", {
    name: "ci-cd-runner-asg",
    desiredCapacity: 2,
    maxSize: 2,
    minSize: 1,
    healthCheckGracePeriod: 300,
    healthCheckType: "EC2",
    launchTemplate: {
        id: launchTemplate.id,
        version: `$Latest`,
    },
    vpcZoneIdentifiers: vpc.privateSubnetIds,
    initialLifecycleHooks: [{
        name: "asg-events-hook",
        defaultResult: "CONTINUE",
        lifecycleTransition: "autoscaling:EC2_INSTANCE_TERMINATING",
        notificationTargetArn: asgEventsTopic.arn
    }]
});

the sns topic is assigned as a

notificationTargetArn

to the ASG. I noticed during delete the topic subscription is deleted before the ASG. If it were the other way around, ASG deleted prior to topic subscription, deleting ASG would fire lifecycle transtion events for the EC2 hosts. In thread, I'm posting the output of

pulumi delete

bug may be / am I missing something?

is it forbidden to use Outputs as resource name? seems like pulumi can't resolve those

hi, I'm looking into Pulumi to set up Azure resources from C# and I'm wondering about one thing: for example, when creating an SQL server, you pass one `name`parameter directly in the constructor, and then there is a property `ServerName`in the

ServerArgs

parameters, too. The same goes for CosmosDb: there is the name parameter of the constructor and the

AccountName

property of

DatabaseAccountArgs

. What's the deal here? Why are there two different ways of specifying the name? What happens if I don't specify the one in the properties? Thank you, Mark @ DEON

hi, still there's no way to clean up unused pulumi plugins? I can't see any related CLI command to figure out which plugins are used by which project/stack, but I'm pretty sure we don't use aws 1.7.0 or cloudflare 1.2.0 etc. plugins anymore. The total plugin cache is almost 7GB, which is unnecessarily big and I want to clean this up, but removing all of the plugins, not a solution, because as far as I know pulumi still does not redownload a plugin if it thinks that already installed and have to do it manually, one by one. Am I correct? What are the alternatives here?

Do I still have to have a CLI if I’m running pulumi commands using Kubernetes operator? I also wanted to use the automation API’s with Kubernetes operator

the documentation for SignalR for typescript shows that one specifies cors so:

cors: [{
        allowedOrigins: ["<http://www.example.com>"],
    }]

but if I do that, there is a com/transpiler error: Type '{ allowedOrigins: string[]; }[]' is not assignable to type 'Input<SignalRCorsSettingsArgs> | undefined'. Type '{ allowedOrigins: string[]; }[]' is missing the following properties from type 'OutputInstance<SignalRCorsSettingsArgs>': apply, getts(2322) signalR.d.ts(118, 5): The expected type comes from property 'cors' which is declared here on type 'SignalRArgs'

so how's the right way?

nevermind, the correct way is:

cors: {

        allowedOrigins: ["http://www.example.com"],     } would be good to adjust the docs

they also talk about `

new azure.signalr.Service

while the current way is

new SignalR

Is there any documentation somewhere on serving private pulumi plugins with the

--server

flag on

pulumi plugin install

? Can the URL be blob storage? Is there a specific directory structure it is expecting? etc.

How does the Azure Native provider code override the sdk generation so that given the name

azure-native

it generates a dotnet project sdk called

Pulumi.AzureNative

instead of

Pulumi.Azure-native

?

Hello guys did any body face "*error:* spawn docker ENOENT" while creating an ECS ?

Hi, I tried to import eks cluster. However, I got these errors.

Computed attributes cannot be set: Computed attributes cannot be set, but a value was set for "vpc_config.0.vpc_id".. Examine values at 'Cluster.VpcConfig.VpcId'.

How can I resolve this ?

Does pulumi have something like https://registry.terraform.io/ yet? Or is searching npm the only way to discover community components (not sure of the correct terminology here)?

Run pulumi/setup-pulumi@v2
  with:
    pulumi-version: 3.7.0
Configured range: 3.7.0

Error: Could not find a version that satisfied the version range

how does one get the connection string for a NamespaceAuthorizationRule for an Azure servicebus from Typescript? the documentation says NamespaceAuthorizationRule should have an output primaryConnectionString, but the d.ts says it has only name, rights and type as outputs.

Hello i use concourse ci/cd to run pulumi.. recently got issue while trying to create ECR. because container i was using did not have prebuilt docker image. are there any containers available which has pre build docker image with pulumi & nodejs ? Container i used was --- /******platform: linux image_resource: type: docker-image source: repository: pulumi/pulumi-nodejs tag: "3.6.1"*******/

The pulumi-provider-boilerplate as it currently exists - if you clone it and run the SDK codegen - seems to have a bug. It generates the dotnet SDK without a package reference to the

Pulumi

NuGet package (see here), so the generated dotnet SDK can't build as it is. Originally I thought OK maybe the stock

schema.json

is just missing an argument to add that reference, but the nodejs SDK generated from the same provider code does have the reference. So I'm wagering it was a bug potentially introduced during changes made to release

azure-native

? Any thoughts on what I need to do so that the codegen adds that reference?

I know Civo is a little bitty player in the scheme of things, so maybe a Civo-specific channel is not justified. However, I kind of hate posting Civo-specific questions into General. I think something like an Other Cloud Providers channel would be warranted for discussions around Civo, DigitalOcean, Linode, etc. Thoughts? Or, maybe I’m just missing a channel that is well-suited for those provider-specific kinds of discussions?

Hi everyone! I am new with Pulumi. I have a trouble when I try to import a Lambda function in AWS. I use

pulumi import aws:lambda/function:Function test_lambda my_test_lambda_function

but, this does not work, error: "Preview failed: handler and runtime must be set when PackageType is Zip" I don't have idea how set this parameters or if have other way to import a Lambda. I hope I have explained my problem well, I appreciate any help. Thanks!

Hello. I’m working on unit testing with Mocha. But when I run my tests with

node_modules/mocha/bin/mocha -r ts-node/register index.ts

, none of the configuration values from my project’s stack are loaded. Any tips would be greatly appreciated.

I’m having an issue where the AWS Provider doesn’t seem to be respecting the

sessionName

param passed in. We have conditions setup on our service accounts that only allow assume role usage if the session name matches the correct phrase. I have Pulumi setup to match this

_sessionName_: 'agent-example'

but in my cloudtrail logs I can clearly see the assume role operation denied and the request parameters show as “null” where it should show the sessionName.

Hello! I'm trying use node-canvas npm module in aws lambda. And I build node modules on Windows, but aws lambda runs on linux and i get error about elf header. How i can fix it in? Any help would be so much appreciated!

how does one upgrade one's account from "individual" to "team"?

Why, when I try and delete a specific resource with the --target property, does Pulumi complain that it can't delete it without deleting the child resource X? If I am telling it to delete the parent resource then I want it to delete all the child resources

how do i run stack programatically without using

pulumi up

?

Is deleting custom resources special in any way? I noticed that If I try to delete one that has children,

--target-dependents

has no effect. I must be missing something here.