While applying some changes to a stack containing an eks cluster, I've hit the following:

stack:Stack::cluster-nodes: UPDATE_ROLLBACK_COMPLETE: ["The requested instance type's architecture (x86_64) does not match the architecture in the manifest for ami-0deb321f7b1c07d4c (arm64). Launching EC2 instance failed."]

The stack usually applies cleanly, and the changes were not directly related to the eks config. I have noted that randomly the ami for nodes in the eks cluster have updated when I've done a

pulumi up

- although a little disconcerting, I haven't worried about it - but now it's selected an invalid ami (it's not clear when / how it looks this up) and won't update. I don't see any way to lock this down to only change when explicitly updated. Has anyone else run into the error above?

is there a way to

await

an Output<>

seeing this as part of diff preview for an ecs service. The task definition being supplied to the service resource is being updated along with the task definition itself. I’d expect to see

output

as a placeholder for the value update, is this normal?

taskDefinition                 : "arn:aws:ecs:eu-west-1:xxxxx:task-definition/xxxxxr:18" => {"Element":{"V":""}}

Is there any way to attach multiple policies to a role with AWS IAM Role Policy Attachment? It seems that adding multiple role policy attachments to a single role is overwriting the previous attachments. Seems like it might be possible with raw Terraform using Count as documented here: https://stackoverflow.com/questions/45486041/how-to-attach-multiple-iam-policies-to-iam-roles-using-terraform but struggling getting it to work with Pulumi

Are there any plans to show parent-child nesting of resources in the

<http://app.pulumi.com|app.pulumi.com>

previews?

The new CloudWatch Dashboard stuff added to awsx in 0.18.2 looks nice! Will there be support for getting

.metrics

from a

FargateService

?

The second s3 bucket arn shows up as

[Object Object]/*

for me, is there any workaround for this?

let rolePolicy = new aws.iam.RolePolicy(`${name}-s3-access`, {
    role: role.id,
    policy: pulumi.output({
      Version: "2012-10-17",
      Statement: [{
        Effect: "Allow",
        Action: [
          "s3:ListBucket",
          "s3:HeadBucket",
          "s3:PutObject",
          "s3:GetObject"
        ],
        Resource: [ bucket.arn, `${bucket.arn}/*` ]
      }]
    }).apply(JSON.stringify)
  }, {
    parent: role
  });

after a week of pain i've finally got my stuff to work, pulumi's awesome again hooray

I'll love to speak to a team member at Pulumi

Are there any examples of installing a helm chart using values from a YAML config? Equivalent to: helm install -n ingress stable/nginx-ingress -f my-config.yaml I found kubernetes.helm.v2.Chart, but wasn't sure how to pass along the YAML file

Hi, in EKS modules, i cannot apply label to default nodegroup

const cluster = new eks.Cluster('staging-auth', {
  nodeAmiId: 'ami-0bfedee6a7845c26d',
  vpcId,
  subnetIds: subnet.split(','),
  deployDashboard: false,
  instanceType: 't2.medium',
  desiredCapacity: 2,
  nodeAssociatePublicIpAddress: false,
  minSize: 2,
  maxSize: 2,
  nodekeyName: 'aws_tokyo',
  storageClasses: 'gp2',
  labels: {
    cluster: 'proxy',
  },
});

is somethings in my conf ?

During a refactoring I implemented a few higher-order components that group some AWS primitives which were part of my main custom component before. So imagine I had the custom component `A `with subordinated components

C¹

,

C²

and

C³

. Now I implemented two components called

B¹

and

B²

which now have

C¹

and

C²

and

C³

as subordinated components respectively and those two components have the parent

A

. When running

pulumi up

the preview shows that this would cause Pulumi to tear down

C¹

,

C²

and

C³

and create them new under

B¹

and

B²

. Since those primitives represent important production infrastructure this would cause a brief service disruption for end users which I would like to avoid. Is there a way to tell Pulumi to just move the components without actually modifying the underlying infrastructure?

Any reason why the EKS security group that manages the communication between the k8s API server and the node instances only allows ports 1025-65535? https://github.com/pulumi/pulumi-eks/blob/master/nodejs/eks/securitygroup.ts#L48

Because of that I got into an issue with accessing rancher's integrated monitoring grafana dashboard: https://github.com/rancher/rancher/issues/19740 .

it would be extra-nice if you could specify a local folder or file:// url to

pulumi new

for the template

Hi! Are there any best practices on how to manage code? I have to admit that I only want to use Pulumi for CI with traditional yaml files, but don't know where to put the yml files. Is it okay to create a new repo, just for the general infrastructure with just yml files? And when I have a specific repo which has some resources (deployment, service) to put that into the respective repo? i.e.

my-project-k8s
  postgres.yml
  ingress.yml
  ...

my-api
   deployment.yml

or is it better to manage all resources in a single

my-project-k8s

repository?

So I just used Pulumi to provision a SSM SecureString parameter, with a value defined by a secret encrypted config value. I was disappointed to see that the value is visible in cleartext in the Pulumi stack. Is there any way to provision a SSM value, preserving its secrecy?

This is probably style as much as anything but what are people doing for k8s deployments to multiple clusters? Are you looping or something in your pulumi program with a new provider per cluster, or are you making a stack per cluster and managing that loop outside of pulumi (shell script or similar)?

Before I embark on this… anybody deploy the Prometheus Operator via Pulumi?

Hey pulumi team, regarding the slowness of updates on certain changes https://github.com/pulumi/pulumi/issues/2674 , I notice that updating to a new container image for a Fargate service, takes about ~8min to finish, though it seems like the actual container starts running much more quickly than pulumi reports success for it. Just some anecdata.

When I want to use Pulumi with Github actions, do I need a paid account?

hello I was asked to join here to discuss an issue I am having with 1.14

I see this error when attempting to do the graceful rollout tutorial with 1.14. Only happens on the second up command, first deployment works just fine.

I ONLY have 1.14 kubectl installed, I verified this multiple times.

and I am using minikube which also is using 1.14

Hey 👋 - I created a Pulumi Provider for Hetzner Cloud (

hcloud

). What would be the release strategy for making it public? 🙂 I saw some deployment scripts in the boilerplate repository which transfer the provider plugin to a Pulumi S3 bucket, etc. I guess this is for the "official providers". What about community providers? 🙂

Hi. My company, Compute Software, recently launched our first SaaS product and pilot program. Our solution dynamically optimizes and automates scaling in the cloud and delivers the greatest business value and efficiency from cloud usage. Would love to show the Pulumi team how we could help optimize scaling. Who would be the right person to talk to? PS: We use Pulumi on a daily basis to deploy all our environments. It has been a lifesaver in getting things up and running quickly 🙂

is there a docker image built off of the master branch? there was PR merged that I have been waiting on and as far as I can tell the docker images are only built off of images pushed to the pulumi s3 bucket