Nov. 17 releases: `dependOn` Helm charts, new Elastic Cloud provider, functions support outputs, set the CLI's default organization

Posted on

Alex Mullans Alex Mullans

It’s been another exciting few weeks here at Pulumi! We’ve caught our breath from Cloud Engineering Summit (don’t forget to check out the talks if you haven’t yet!) and we’re back to adding new value and highly-requested fixes across the Pulumi Cloud Engineering Platform. Read on to learn about new providers, new enhancements to the core Pulumi experience, and more!

Pulumi Registry, Pulumi Packages, & integrations

dependsOn can now await a fully-deployed Helm Chart resource

Pulumi’s dependsOn option enables you to write programs that deploy resources in a specific order. Previously, using this option with a Helm Chart resource could be confusing because Pulumi would consider the resource ready once it had registered the chart, even if Helm was still deploying some of the resources defined by the chart. Now, you can depend on a new ready attribute that will ensure that none of the dependent resources are deployed until the chart is deployed. For example, the code sample below creates an NGINX Ingress Controller and then creates a Kubernetes ConfigMap that depends on the controller being fully deployed.

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

const nginxIngress = new k8s.helm.v3.Chart("nginx-ingress", {
    chart: "nginx-ingress",
    version: "1.24.4",
    namespace: "test-namespace",
    fetchOpts:{
        repo: "https://charts.helm.sh/stable",
    },
});

// Create a ConfigMap depending on the Chart. The ConfigMap will not be created until after all of the Chart
// resources are ready. Note the use of the `ready` attribute; depending on the Chart resource directly will not work.
new k8s.core.v1.ConfigMap("foo", {
    metadata: { namespace: namespaceName },
    data: {foo: "bar"}
}, {dependsOn: nginxIngress.ready})
This change impacts the older Helm Chart resource. The new Helm Release resource always waits for the Helm chart to be fully deployed.

Learn more about dependsOn and Helm charts in this GitHub issue

New provider: Elastic Cloud

You can now provision Elastic Cloud resources like ElasticSearch using Pulumi! Visit the ElasticCloud (EC) provider on Pulumi Registry to get started. The example below shows how you can create an Elastic Cloud deployment in just a few lines of code:

const ec = require("@pulumi/ec")

const REGION = "us-east-1";

const latestVersion = elasticCloud.getStack({
    region: REGION,
    versionRegex: "latest"
});

new elasticCloud.Deployment('my-deployment', {
    region: REGION,
    version: latestVersion.then((x: { version: string; }) => x.version),
    deploymentTemplateId: "aws-io-optimized-v2",
    elasticsearch: {}
});

import * as elasticCloud from "@pulumi/ec";

const REGION = "us-east-1";

const latestVersion = elasticCloud.getStack({
    region: REGION,
    versionRegex: "latest"
});

new elasticCloud.Deployment('my-deployment', {
    region: REGION,
    version: latestVersion.then((x: { version: string; }) => x.version),
    deploymentTemplateId: "aws-io-optimized-v2",
    elasticsearch: {}
});

import pulumi_ec as ec

REGION = "us-east-1"

latest_version = ec.get_stack(
    region=REGION, version_regex="latest").version

deployment = ec.Deployment('my-deployment', region=REGION, version=latest_version,
                           deployment_template_id="aws-io-optimized-v2", elasticsearch={})

import (
  "github.com/pulumi/pulumi/sdk/v3/go/pulumi"
  ec "github.com/pulumi/pulumi-ec/sdk/go/ec"
)

func main() {
  pulumi.Run(func(ctx *pulumi.Context) error {
        version, err := ec.GetStack(ctx, &GetStackArgs{
      Region:       pulumi.String("us-east-1"),
      VersionRegex: "latest",
        }
    if err != nil {
      return err
    }

    _, err := ec.NewDeployment(ctx, "my-deployment", &ec.DeploymentArgs{
            Region:               pulumi.String("us-east-1"),
      Version:              version.Version,
            DeploymentTemplateId: pulumi.String("aws-io-optimized-v2")
      ElasticSearch:        &ec.DeploymentElasticsearch{},
    })
    if err != nil {
      return err
    }

    return nil
  })
}

using System.Collections.Generic;
using System.Threading.Tasks;
using Pulumi;
using ElasticCloud = Pulumi.ElasticCloud;

class Program
{
    static Task Main() =>
        Deployment.Run(() => {
            var latest = Output.Create(ElasticCloud.GetStack.InvokeAsync(new ElasticCloud.GetStackArgs
            {
                Region = "us-east-1",
                VersionRegex = "latest",
            });

            var deployment = ElasticCloud.Deployment("my-deployment", new ElasticCloud.DeploymentArgs
            {
                Region = "us-east-1",
                Version = latest.Version,
                DeploymentTemplateId = "aws-io-optimized-v2",
                ElasticSearch = {},
            });
        });
}

AWS Native provider supports auto-naming

Most Pulumi providers support auto-naming for resources, which makes it easier to have multiple stacks of the same Pulumi program by avoiding naming conflicts, among other benefits. Now, our AWS Native provider joins the club with full auto-naming support.

Learn more about AWS auto-naming in this GitHub issue

New resources in the AWS Native provider

We shipped new versions of the AWS Native provider (0.3.0 through 0.4.0) that added 38 new resources giving you access to resources like AWS::EKS::Cluster, AWS::Lightsail::Database, and several new resources in the AWS::Redshift namespace.

New resources in the Azure Native provider

We shipped new versions of the Azure Native provider (1.43.0 through 1.45.0) that collectively added 4 new resources giving you access to resources like Cognitive Services.

Pulumi CLI and core technologies

In this milestone, we shipped Pulumi version 3.17.0. The full list of changes in this version is available in the changelog; read on to learn about some of the biggest changes.

Functions now support outputs

When creating complex infrastructure, it’s common to pass information about one resource to a function provided by another resource. Previously, doing this in Pulumi required you to use apply to get the information from the first resource and pass it into the function. Now, we’ve simplified the process so you can reference outputs directly from the function—no apply needed. To take advantage of these changes, make sure you’ve updated to the latest versions of Pulumi and the providers you’re using.

Learn more in the “Functions now accept outputs” blog post

--json flag for pulumi {update|destroy|preview|refresh}

When automating or troubleshooting Pulumi operations, it can be helpful to see all of the details emitted by the Pulumi engine rather than the standard progress bar. Now, you can! Just add --json to your pulumi {update|destroy|preview|refresh} command.

Learn more about the new --json flag in this GitHub issue

Python 3.6 support will be deprecated soon

Python 3.6 will be end-of-lifed (“EOLed”) on December 23, 2021. Around that time, Pulumi will drop support for Python 3.6. If you’re using Python 3.6, consider upgrading to a newer version of Python. The Pulumi CLI will display a message when it runs a Pulumi program using Python 3.6.

Learn more about Python 3.6 deprecation in this GitHub issue

Pulumi Service & Pulumi.com

Set the default organization for newly-created stacks

If you’re using Pulumi as part of a team or company, rather than as part of a personal project, you likely want to ensure that any infrastructure you deploy is registered in your company’s Pulumi organization, rather than in your personal Pulumi account. Previously, you needed to either create each stack inside the organization (pulumi new -s <org>/<stackname>) or transfer stacks into the organization once created. Now, you can set the default organization for Pulumi to use when creating new stacks: pulumi org set-default <org>. Watch the asciinema recording below to get a feel for how it works:

Learn more about setting the default organization in this GitHub issue

Share this post

PulumiUP - September 18, 2024. Register Now.