set the channel description: Using Kubernetes with Pulumi

Alright, let’s get the new year started with an update on one of our oldest and most upvoted k8s issues: https://github.com/pulumi/pulumi-kubernetes/issues/29 I’ve got a proof of concept PR up to address this use case: https://github.com/pulumi/pulumi-kubernetes/pull/936 Interested in any early feedback on the approach, particularly from anybody using tools other than Pulumi (kubectl, Argo, Skaffold, etc.) to deploy k8s apps.

I might try to use that for rendering 10+ helm charts and using the YML in pulumi , instead of the helm chart resource. Helm charts add quite a bit of time to

preview/update

Filed this issue: https://github.com/pulumi/pulumi-kubernetesx/issues/43

Another issue: https://github.com/pulumi/pulumi-kubernetesx/issues/44

I have a a stable helm chart (airflow) https://github.com/helm/charts/blob/master/stable/airflow/values.yaml. I am able to override values in pulumi code when the values.yaml has a map, array, string etc. How can I override affinity which needs new line and indentation ?

What is a cleaner way to specify dependencies between stacks? Exporting objects often means the CLI logs these large objects (I know there is a flag to hide that though).

.

How would one go about importing a kubernetesyamlConfigFile as all the resources are dependents?

Hi guys , For some helm charts , you will need to apply first your CRD definition object than your crd instance. I’m wondering if there is a way to indicate that via Pulumi . Thanks #kubernetes #helm

Anyone else getting this issue? Not sure if its related to the latest provider release We don’t use implicit provider

kubeconfig

anywhere, only grabbing from

eks.Cluster.provider

Error: invocation of kubernetes:yaml:decode returned an error: unable to load Kubernetes client configuration from kubeconfig file: invalid configuration: no configuration has been provided
        at monitor.invoke (/root/project/node_modules/@pulumi/pulumi/runtime/invoke.js:172:33)
        at Object.onReceiveStatus (/root/project/node_modules/grpc/src/client_interceptors.js:1210:9)
        at InterceptingListener._callNext (/root/project/node_modules/grpc/src/client_interceptors.js:568:42)
        at InterceptingListener.onReceiveStatus (/root/project/node_modules/grpc/src/client_interceptors.js:618:8)
        at callback (/root/project/node_modules/grpc/src/client_interceptors.js:847:24)

When I use

kubernetesx.PodBuilder(...).asStatefulSetSpec()

, is there support to get my

volumeClaimTemplates

in the spec?

I can’t see any way to add it based on the code: https://github.com/pulumi/pulumi-kubernetesx/blob/master/nodejs/kubernetesx/kx.ts#L185-L197

i create an aks cluster, and then use the following to create a k8sprovider which is then passed into each kubernetes resource using the

provider

argument of

CustomResourceOptions

.

const k8sProvider = new k8s.Provider('k8sProvider', {
  kubeconfig: k8sCluster.kubeConfigRaw
})

This has been working fine, until i made a config change on the aks cluster where the cluster needed to be destroyed and recreated. The new aks instance was created successfully; however, the kubernetes resources were not updated to use the new provider. I have dealt with this in the past by exporting the stack config, manually editing all of the resources to use the new provider, then importing the stack config and running pulumi up. Am I doing something wrong in how i specify the provider on the resources, or is this a limitation of pulumi?

We are using pulumi with typescript and a pulumi yaml config file. We have included “enableIstio”: false using “pulumi config set enableIstio false”. If we use let blabla = config.getBoolean(“enableIstio”) || true it always results in true so I assume we cannot use this example? I know that typescript defines it as such. For let blabla = config.getNumber(“somenumber”) || 20 is working correctly except in the case where your number is a 0 (=false). A bit of a gotcha as a lot of these code examples use || and not ??

how does pulumi handle rollback in kubernetes? specifically situations such as configmap linked to applications.. if previous upgrade was happened only to configmap, and not to application(kind: deployment).. during current upgrade if i encounter any issue, i can only rollback app (deployment), but can't get previous conifgmap.. Currently helm handles this with chart release history, wondering if pulumi can handle too

i see pulumi supports history, but I don't find a way to apply specific version if needed

hey, I don't see data sources in pulumi-kubernetes... is there any way to refer to non-managed k8s resources? I want to have service account token as output and it is inside secret which is created by k8s for the

ServiceAccount

@elegant-dress-88912 you may want to try following? https://www.pulumi.com/docs/intro/concepts/programming-model/#programmatically-creating-secrets

@orange-ghost-99337 this won't work. I need data source for the secret, like in terraform: https://www.terraform.io/docs/providers/kubernetes/d/secret.html

got it.. i mistaken your question..:)

Any suggestion around - https://github.com/kubernetes-sigs/aws-alb-ingress-controller/issues/1137

hi - i'm using cert-manager and deploying it via yaml.ConfigFile but then when i try and configure a DNS issuer the cluster immediately rejects it because the webhook isn't ready yet... i can't find a .status field or anything else that looks relevant to poll / wait against has anybody solved this or got any suggestions how I can avoid this race condition please?

Hi all! Just curious on the timeline for helm v3 support. Thanks!

Hi, is there any reason why

yaml.ConfigGroup

yaml

field does not accept

Pulumi.Input<string>

? This way I'm unable to pass a string created from an output, for example from a secret

hi, is it possible to delete some resource from helm chart during transformation or any other way?

is anybody here using kubernetesx for production purposes? I’m going to import a bunch of k8s resources into pulumi pretty soon and kx looks really enticing, so I’m wondering how stable it is atm

I’m willing to make some changes along the way of course, I know it’s not completely stable

but like, are the APIs mostly set?

We're using Pulumi to build a Kubernetes cluster on GCP, with the Istio integration enabled (https://cloud.google.com/istio/docs/istio-on-gke/overview). Istio on GKE creates all the necessary k8s resources to run Istio, we do not have a reference to them in Pulumi. In order to setup TLS and configure domains, the docs recommend us to edit the existing resources. We can do this with Pulumi via an import, but (if I've understood correctly) this means we cannot build our setup with a single

pulumi up

. Instead, we need one

pulumi up

where the

import

simply reflects the existing resource, and then another one where we edit the code to change the resource. We've verified this multi-stage flow works, but it feels clunky (currently we have to manually edit the code between the first and second pulumi up). Are there any tricks we can use to combine into a single stage? One thought was that maybe there is a variable that denotes whether or not the resource is already referenced. Then we could run pulumi twice, and use a switch statement to distinguish the two cases. Feels very hacky though.