Is there a way when creating a service to tell it to not wait for any pods to be ready? Specifically, we create our services prior to creating the Deployments/DaemonSets, as we depend on the service environment variable injection to happen into the later-created pods. With Pulumi, I’ve expressed the dependency (i.e. the Deployment has a DependsOn the Service objects), but the Service itself never becomes ready as it is in stuck in the “Finding Pods to direct traffic to” stage (1/3). I’d like to tell it to not do that.

This is more of a pattern question, but what would be a good way to do something like: 1. Create a namespace 2. Instantiate a provider that is defaulted to that namespace 3. Use that provider for all subsequent resources

Also, separately, I found what are likely two small bugs in the

kube2pulumi

online site - should I just open GitHub issues for them?

hello Pulumi team, I updated my ticket with conclusive findings on a bug with secret state encryption causing

pulumi preview --diff

to not print out anything useful

Is there a good way to handle “dependencies” within a Helm chart deployed via Pulumi? In this specific case, I am deploying the Superset Helm chart, and it has some implicit dependencies between some ConfigMap objects and Jobs/Deployments, so if I update the ConfigMap (via the passed in

values

), things don’t get re-run. In a normal setting I’d just explicitly declare the dependencies (or have it be implicit via the resources), but the Helm chart is a single unit, and not clear to me if I can do it via a Transformation.

I miss the fast-feedback of TDD workflows in pulumi land, and I remember seeing some test driven examples (I think compliance-related, ensuring some resources are of a specific type, kind of stuff) Is anyone building their stacks in a test-driven way? I'm planning to refactor one stack into two, and I'd like an automatic test to tell me that everything's green. I believe I could use the outputs/exports to have the stacks inter-communicate(?) What does "test-driven" mean in a cloud engineering context anyway? I'm thinking I should be able to boot-up an Ubuntu VM with microk8s in it, boot up the k8s pulumi stack, run some tests, and then tear everything down. Later this could fit into some CI pipeline. (I'm working locally, not on a public cloud, and the feedback loop is painfully slow)

for some reason, my core.v1.Secret resource wants to be replaced on every pulumi run arguing that "data" has changed (it has not) - how do I debug this?

Is this the best way to get the kubeconfig from an EKS cluster in Go to pass to another provider? https://github.com/pulumi/examples/blob/e3b4e1734fc827451e3bccf623468c320015a084/aws-go-eks/main.go#L143

hi. in here https://www.pulumi.com/docs/guides/adopting/from_kubernetes/ it shows

import * as k8s from "@pulumi/kubernetes";

// Deploy the latest version of the stable/wordpress chart.
const wordpress = new k8s.helm.v3.Chart("wpdev", {
    repo: "stable",
    chart: "wordpress",
    version: "9.0.3",
});

// Export the public IP for WordPress.
const frontend = wordpress.getResource("v1/Service", "wpdev-wordpress");
export const frontendIp = frontend.status.loadBalancer.ingress[0].ip;

Is there a way to get this code to run as is or must it be modified with apply like so

import * as k8s from "@pulumi/kubernetes";

// Deploy the latest version of the stable/wordpress chart.
const wordpress = new k8s.helm.v3.Chart("wpdev", {
    repo: "stable",
    chart: "wordpress",
    version: "9.0.3",
});

// Export the public IP for WordPress.
const frontend = wordpress.apply(wordpress => wordpress.getResource("v1/Service", "wpdev-wordpress"));
export const frontendIp = frontend.apply(frontend=>frontend.status.loadBalancer.ingress[0].ip);

Thanks

So, I accidentally deleted a deployment managed by Pulumi

what steps do I have to take to reconcile the state?

I tried

pulumi refresh

and that seems to have removed these resources from the state, but now I can’t

pulumi up

because there are still other resources (eg. PVCs) still hanging around

I’ve read https://gist.github.com/clstokes/62d8ff84ada8cd414a88d4f1030f5ddc on resource adoption but it seems counter-intuitive for me to change my stack to look up the resources, then revert it

What’s the right way to work with

pulumi.NewFileAsset

if I want to use it as an input to

StringData

for a secret?

Is there an annotation or something I can use to force recreation of some resources?

I found that I had a deployed statefulset with a container that wouldn’t start up (because I’d messed up a volume name) and when I tried to redeploy the fix, it was just waiting for the pod to go away

Anybody know how to import these custom resources (like these above)? I'm looking to just do them one by one on the CLI no code. I can import standard resources, but I'm not sure what the syntax is for the CLI for custom resources.

i am trying to deploy

<https://github.com/kubernetes-sigs/aws-efs-csi-driver/tree/master/charts/aws-efs-csi-driver>

as a pulumi_kubernetes.helm.v3.Chart but it fails because

<http://csidrivers.storage.k8s.io|csidrivers.storage.k8s.io> "<http://efs.csi.aws.com|efs.csi.aws.com>" already exists

…. this happens because the chart includes a pre-install helm hook to remove the existing efs-csi driver and pulumi doesnt support helm-hooks (

<https://github.com/pulumi/pulumi-kubernetes/issues/555>

) . . . what is the best way for me to work around this ?

What’s the typical way of renaming a resource? Specifically, let’s say I have:

appsv1.NewDeployment(ctx, "myDeployment", &appsv1.DeploymentArgs{

and I want to change it to

appsv1.NewDeployment(ctx, "my-deployment", &appsv1.DeploymentArgs{

If I do that, it fails because it says the Deployment already exists with that name (note that I have given it an explicit name, so it is not using autonaming)

Not sure if this is the right place to ask. I have EKS setup with pulumi, but I would like to deploy a set of infrastructure application as part of the cluster configuration. Instead of install each individual helm charts using pulumi helm chart provider, I am trying to deploy them using Helmfile. Is there any way that I can integrate helmfile into pulumi workflow, say I want to automatically trigger "helmfile sync" as part of "pulumi up" and "helmfile destroy" as part of "pulumi destroy"? Has anybody done this before? How do you do it?

any ideas why I am getting this error?

error: no resource plugin 'kubernetes-v1.1.0' found in the workspace or on your $PATH, install the plugin using `pulumi plugin install resource kubernetes v1.1.0`

Hi! I have a helm repo in my S3 storage. I want to apply this helm charts with k8s.helm.v3.chart but next error appear.

Error: invocation of kubernetes:helm:template returned an error: failed to generate YAML for specified Helm chart: failed to pull chart: could not find protocol handler for: s3

My cluster is in 1.16 ...

Hey all, I am having trouble deploying ingresses in EKS using the alb aws load balancer controller

<https://aws.github.io/eks-charts>

The error that I am getting looks like this:

i notice that on every pulumi up, pulumi deletes and creates a new secret for the alb chart’s tls cert

i don’t know why in the first place pulumi is doing this. There are no changes to the alb chart i have deployed. The only changes that go on are deploying ingresses