I just used pulumi deployments for the first time - thanks to @red-match-15116 for the heads up on getting that configured in my stack settings. 🎉

Question on pulumi deployment of a .net 7 app - is there a runner image I can use that supports .net 7? I understand I need to use a custom executor image.

With the pulumi CLI, pulumi up gives informative output on plugin installs e.g.:

Downloading plugin: 158.80 MiB / 158.80 MiB [====================] 100.00% XmYs
[resource plugin aws-5.40.0] installing

However, when using the automation api with

optup.ProgressStreams

, I'm only getting this output:

@ updating.....................................................................

for as long as it takes plugins to install. Is there a way to get the same more verbose output with the automation API? Thanks!

I have a dotnet application that using Automation API to execute Local Program. I run it on my local windows machine it is fine, but I run it on a docker linux container, it throws error message. I can confirm that the folder exists with correct permission, and I can see Pulumi has added Pulumi.yaml file to this folder. Any help will be appreciated.

"errorMessage": "An error occurred trying to start process 'pulumi' with working directory '/tmp/automation-33ihygga.ijh'. No such file or directory",
"exceptionType": "Win32Exception"

at System.Diagnostics.Process.StartCore(ProcessStartInfo startInfo)
   at System.Diagnostics.Process.Start()
   at CliWrap.Utils.ProcessEx.Start() in /home/runner/work/CliWrap/CliWrap/CliWrap/Utils/ProcessEx.cs:line 65
   at CliWrap.Command.ExecuteAsync(ProcessEx process, CancellationToken cancellationToken) in /home/runner/work/CliWrap/CliWrap/CliWrap/Command.cs:line 372
   at Pulumi.Automation.Commands.LocalPulumiCmd.RunAsyncInner(IList`1 args, String workingDir, IDictionary`2 additionalEnv, Action`1 onStandardOutput, Action`1 onStandardError, EventLogFile eventLogFile, CancellationToken cancellationToken)
   at Pulumi.Automation.Commands.LocalPulumiCmd.RunAsync(IList`1 args, String workingDir, IDictionary`2 additionalEnv, Action`1 onStandardOutput, Action`1 onStandardError, Action`1 onEngineEvent, CancellationToken cancellationToken)
   at Pulumi.Automation.LocalWorkspace.PopulatePulumiVersionAsync(CancellationToken cancellationToken)
   at Pulumi.Automation.LocalWorkspace.CreateStackHelperAsync(InlineProgramArgs args, Func`4 initFunc, CancellationToken cancellationToken)

We’re trying to use the automation api with nextjs but are getting complaints about multiple pulumi versions running, but there is only one version based on yarn why. Is https://github.com/pulumi/pulumi/issues/5449 still an issue given that it was closed as resolved?

Hello 👋 hoping to get a comment from the Pulumi team on this one... I've just started seeing this error pop up out of nowhere (pretty sure I've not changed the code at all since I last ran it)

/Users/REDACTED/code/infrastructure/node_modules/@grpc/grpc-js/src/call.ts:81
  const error = new Error(message);
                ^
Error: 14 UNAVAILABLE: No connection established
    at callErrorFromStatus (/Users/REDACTED/code/infrastructure/node_modules/@grpc/grpc-js/src/call.ts:81:17)
    at Object.onReceiveStatus (/Users/REDACTED/code/infrastructure/node_modules/@grpc/grpc-js/src/client.ts:356:55)
    at Object.onReceiveStatus (/Users/REDACTED/code/infrastructure/node_modules/@grpc/grpc-js/src/client-interceptors.ts:455:34)
    at Object.onReceiveStatus (/Users/REDACTED/code/infrastructure/node_modules/@grpc/grpc-js/src/client-interceptors.ts:417:48)
    at /Users/REDACTED/code/infrastructure/node_modules/@grpc/grpc-js/src/resolving-call.ts:111:24
    at processTicksAndRejections (node:internal/process/task_queues:78:11)
for call at
    at ServiceClientImpl.makeUnaryRequest (/Users/REDACTED/code/infrastructure/node_modules/@grpc/grpc-js/src/client.ts:326:30)
    at ServiceClientImpl.registerResourceOutputs (/Users/REDACTED/code/infrastructure/node_modules/@grpc/grpc-js/src/make-client.ts:189:15)
    at /Users/REDACTED/code/infrastructure/node_modules/@pulumi/runtime/resource.ts:971:42
    at new Promise (<anonymous>)
    at Object.<anonymous> (/Users/REDACTED/code/infrastructure/node_modules/@pulumi/runtime/resource.ts:970:21)
    at Generator.next (<anonymous>)
    at fulfilled (/Users/REDACTED/code/infrastructure/node_modules/@pulumi/pulumi/runtime/resource.js:18:58)
    at processTicksAndRejections (node:internal/process/task_queues:96:5)
error Command failed with exit code 1.

Hi! When using the automation api, where do plugins get installed to? e.g., running this snippet

stack.workspace.installPlugin("aws", "v4.0.0");

I am considering using the automation api to manage resources I am provisioning on behalf of my users. I have pulumi stack code written and parameterized in the way I need. The thing is, my application is a JVM app. As such, I’m pondering how to get this whole setup working in a prod scenario. My current thinking is to bundle my pulumi stack code into a single js executable (use webpack or ncc), package that into an executable binary (e.g., via pkg), add that binary to my container, and, finally, add the pulumi cli to my container. I can then execute pulumi stack operations by interactive with the binary executable I produced. Does this sound like a typical workflow for an automation api user?

I’m taking a look at creating a deployment via the rest api. It mentions a stack must exist prior to creating the deployment. Reasonable. However, looking at the regular rest api for stacks, I do not see any way to create a stack via the rest api. Is this actually the case? So even using the deployment api would still require wrapping the cli to create a stack?

if I have api keys exported as secrets from my main stack (not deployed via automation api), is there a way to grab these and configure a

stack.setConfig

call inside automation api?

I guess this is not currently possible: https://github.com/pulumi/pulumi/issues/9112 Hopefully I can find a workaround

@icy-controller-6092 Are you able to share a few more details around what you’re trying to accomplish? We might be able to help you find a way to do it.

Create secrets in shared stack, export the outputs the IDs for secrets, use stack references to get their values.

Hey there, that's pretty much it. I create workspaces with various cloud platforms in my main stack and export e.g. api keys, then need to configure the providers in the other standalone stacks

Very relevant to many of you building developer platforms and automation. We just launched a feature called Review Stacks 🚀 https://pulumi-community.slack.com/archives/CB36DSVSA/p1686759675815729

Tried to run my first deployment on a remote fly.io server, but I was using the default image with only 256mb of ram, so it ran out of memory and crashed 😅 scaled up to 1gb and that seemed to have done the trick, but may need to now put some kind of queue in place

Very relevant to many of the folks here. #pulumi-deployments now supports automated downstream updates across dependent stacks. Makes it much easier to keep stack references up to date https://pulumi-community.slack.com/archives/CB36DSVSA/p1687280202924809

Short question: for Python, is it possible to have multiple concurrent Pulumi sessions if running them in separate threads? (We currently run a larger set of independent deployment jobs in serial, but would like to run them in parallel if possible...)

Hello, using the automation API I'm trying to handle a resource.get and catch it for when it doesn't exist. I've gotten it to process the except block but the program then stopped building the stack, code below:

try:

user = Output(ccloud.get_user_output(email=config['email'])).apply(lambda user: user)

id = IdHolder(user.id)

return id

except:

inv = ccloud.Invitation(resource_name=name, email=config['email'], opts=ResourceOptions(retain_on_delete=True))

id = IdHolder(inv.users[0].id)

return id

The use case is, see if a user exists if not send an invite. Wrapping the .get_user_output() call with a Output() finally lets the except block execute, creating the invitation, but then the program stops no other resources are created

Hello! This seems like a basic question but I can't seem to find the answer. Is there a way to pass a custom UserAgent parameter to Pulumi in AWS (s.t. the UserAgent listed in CloudTrail will be my custom UserAgent)? Thanks! I'm looking for something like tfexec's

SetAppendUserAgent

https://pkg.go.dev/github.com/hashicorp/terraform-exec/tfexec#Terraform.SetAppendUserAgent

Can I trigger multiple automation programs from a single process? Will there be any issue if the working directory is the same, if I run the programs in parallel?

Are there plans for RemoteStacks to support setting configs?

Hello, I'm attempting to create a Dynamic Resource Provider resource using the typescript automation api from within a vscode extension (vscode:1.80.1, Node.js: 16.17.1) and am receiving the following error. Standard Resource Provider resources work fine without error and running the same dynamic resource code outside of vscode works fine. My vscode extension for testing is here: https://github.com/korosccj/pulumi-test Any help/guidance would be appreciated.

Pulumi Test Extension Running
extensionHostProcess.js:105
Updating (pulumi-test):

extensionHostProcess.js:105
    pulumi:pulumi:Stack test-pulumi-test running 

extensionHostProcess.js:105
    pulumi:pulumi:Stack test-pulumi-test running error: Unhandled exception: Error: Cannot find module 'prototype'

extensionHostProcess.js:105
@ updating....
extensionHostProcess.js:105
rejected promise not handled within 1 second: Error: Cannot find module 'prototype'
Require stack:
- /Applications/Visual Studio <http://Code.app/Contents/Resources/app/out/bootstrap-amd.js|Code.app/Contents/Resources/app/out/bootstrap-amd.js>
- /Applications/Visual Studio <http://Code.app/Contents/Resources/app/out/bootstrap-fork.js|Code.app/Contents/Resources/app/out/bootstrap-fork.js>
- 
extensionHostProcess.js:105
stack trace: Error: Cannot find module 'prototype'
Require stack:
- /Applications/Visual Studio <http://Code.app/Contents/Resources/app/out/bootstrap-amd.js|Code.app/Contents/Resources/app/out/bootstrap-amd.js>
- /Applications/Visual Studio <http://Code.app/Contents/Resources/app/out/bootstrap-fork.js|Code.app/Contents/Resources/app/out/bootstrap-fork.js>
- 
    at Module._resolveFilename (node:internal/modules/cjs/loader:1010:15)
    at Function.s._resolveFilename (node:electron/js2c/utility_init:2:2740)
    at Module._load (node:internal/modules/cjs/loader:858:27)
    at Function.f._load (node:electron/js2c/asar_bundle:2:13330)
    at Function.l._load (/Applications/Visual Studio <http://Code.app/Contents/Resources/app/out/vs/workbench/api/node/extensionHostProcess.js:127:28070|Code.app/Contents/Resources/app/out/vs/workbench/api/node/extensionHostProcess.js:127:28070>)
    at Function.p._load (/Applications/Visual Studio <http://Code.app/Contents/Resources/app/out/vs/workbench/api/node/extensionHostProcess.js:127:25404|Code.app/Contents/Resources/app/out/vs/workbench/api/node/extensionHostProcess.js:127:25404>)
    at Function.u._load (/Applications/Visual Studio <http://Code.app/Contents/Resources/app/out/vs/workbench/api/node/extensionHostProcess.js:91:22191|Code.app/Contents/Resources/app/out/vs/workbench/api/node/extensionHostProcess.js:91:22191>)
    at Module.require (node:internal/modules/cjs/loader:1082:19)
    at require (node:internal/modules/cjs/helpers:102:18)
    at Object.get (/Applications/Visual Studio <http://Code.app/Contents/Resources/app/out/bootstrap-amd.js:6:87|Code.app/Contents/Resources/app/out/bootstrap-amd.js:6:87>)
    at pulumi-test/node_modules/@pulumi/runtime/closure/createClosure.ts:324:49
    at Generator.next (<anonymous>)
    at fulfilled (pulumi-test/node_modules/@pulumi/pulumi/runtime/closure/createClosure.js:18:58)
    at processTicksAndRejections (node:internal/process/task_queues:96:5)

It appears to be an issue when createClosure.addGlobalInfoAsync is serializing the globals and attempts to access the protocol key on a vscode global "_VSCODE_NODE_MODULES", which Proxies the get lookup to

require

#pulumi:createClosure.js
async function addGlobalInfoAsync(key: string) {
    const globalObj = (<any>global)[key];
    const text = utils.isLegalMemberName(key) ? `global.${key}` : `global["${key}"]`;

    if (globalObj !== undefined && globalObj !== null) {
        await addEntriesAsync(globalObj, text);
        await addEntriesAsync(Object.getPrototypeOf(globalObj), `Object.getPrototypeOf(${text})`);
        await addEntriesAsync(globalObj.prototype, `${text}.prototype`);
    }
}

#vscode:bootstram-amd.js
const nodeRequire = require;
if (globalThis._VSCODE_NODE_MODULES = new Proxy(Object.create(null), {
        get: (e, r) => nodeRequire(String(r))
    }), globalThis._VSCODE_PRODUCT_JSON = require("../product.json"), process.env.VSCODE_DEV) try {
    const e = require("../product.overrides.json");
    globalThis._VSCODE_PRODUCT_JSON = Object.assign(globalThis._VSCODE_PRODUCT_JSON, e)
} catch {}

Hi All! For some reason there is a python automation api code that is not working for me.

class PulumiProgram:
    def __init__(self):
        self._work_dir = os.path.join(os.path.dirname(__file__))
        self._data = self._load_config_file('defaults.yaml')
        self._stack = auto.create_or_select_stack('dev', project_name='test-eks', work_dir=self._work_dir,
                                                  program=self._create_resources())
        self._stack.set_config('aws:region', auto.ConfigValue('us-east-1'))
        self._stack.set_config('values', auto.ConfigValue(json.dumps(self._data.get('values'))))

    def preview(self):
        self._stack.refresh(on_output=print)
        self._stack.preview(on_output=print)

    def up(self):
        self._stack.refresh(on_output=print)
        self._stack.up(on_output=print)

    def _load_config_file(self, path: str):
        with open(path, "r") as f:
            data = yaml.full_load(f)
            return data

    def _create_resources(self):
        chart_values = self._data.get('values')
        kube = open("/Users/nathanell/.kube/lab-okta-playground-eks").read()
        k8s_provider = k8s.Provider(resource_name="k8s-provider", kubeconfig=pulumi.Output.secret(kube))
        ui_app_chart = Release(
            "ui-app-micro-fe",
            ReleaseArgs(
                name='nginx-test-nati',
                chart="../helm",
                values=chart_values,
                namespace=f'{chart_values.get("project")}-{chart_values.get("app")}',
                create_namespace=True,
                version='1.0.0'
            ),
            opts=ResourceOptions(provider=k8s_provider)
        )

if __name__ == "__main__":
    PulumiProgram().up()

While the same code is working for me while using Pulumi CLI program:

config = pulumi.Config()
 chart_values = json.loads(config.get('values'))
 kube = open("/Users/nathanell/.kube/lab-okta-playground-eks").read()
 provider = k8s.Provider(resource_name="k8s-provider", kubeconfig=pulumi.Output.secret(kube))
 print(chart_values)
 nginx_chart = Release(
     "nginx-ingress",
     ReleaseArgs(
         name='nginx-test-nati',
         chart="../helm",
         values=chart_values,
         namespace=f'{chart_values.get("project")}-{chart_values.get("app")}',
         create_namespace=True,
         version='1.0.0'
     ),
     opts=ResourceOptions(provider=provider)
 )

This is the error that I'm getting:

kubernetes:core/v1:Namespace (ui-app-nginx):
    error: configured Kubernetes cluster is unreachable: unable to load schema information from the API server: the server has asked for the client to provide credentials

Although I have verified that the kubeconfig is ok and working in the pulumi CLI program Someone can assist?

Hey team! I wonder if it's possible to configure an auto deployment using code instead of going over every stack manually 🤔

Hi We have a pulumi automation API program that should deploy Helm charts to a cluster using pulumi. In order to deploy some chart to a cluster, we need to create a specific kubeconfig file. To do that, we are using a Role-Based kubeconfig file that should create a connection of a specific TTL. This is the code: this is the code that generates the kubeconfig file: Then we are fetching the cluster data and the cluster auth token to pass to the kbueconfig file generator and to get a dynamic kubeconfig file as a provider: The kubeconfig file is working correctly while running pulumi.up() option but while trying to refresh the stack, we are getting an error that tells that the kubeconfig is not updated and we can't communicate with the cluster. This is an issue that was opened as well in the github of pulumi_kubernetes. https://github.com/pulumi/pulumi-kubernetes/issues/2377 Thanks in advance.

Hi there, I’m facing this issue when deployment some snowflake resources, yesterday when running a up my process didn’t finish, so I’m think if it is not a state thing. In addition, the file formats are in Snowflake and also in the state file. I tried dropping it, but it is stuck in this issue.

I guess the reason why the user field here is populated with my name is because it’s my API-key that is used. However, this was pushed by someone else. Is there a way with Automation API that we can set the actor / User somehow?

Would that make sense to add, if not?

Hi there! Am I correct that it's not currently allowed to pass in the path to the policy pack as a parameter for remote stacks? Like this:

const stack = await RemoteWorkspace.selectStack({url: <git repo>})
await stack.up({
    onOutput: <http://console.info|console.info>,
    policyPacks: ['../policy_folder']})