Hi guys, good evening! I'm trying to import an existing resource to my pulumi program but I'm not understanding what the "id" parameter means. As an example, I'm trying to import an iam Role with "aws.iam.Role.get()". The expected parameters are documented as: name – The unique name of the resulting resource. id – The unique provider ID of the resource to lookup. state – Any extra arguments used during the lookup. opts – Optional settings to control the behavior of the CustomResource. How can I get the ID?? Updating: based on a PR thread on pulumi docs and this link, I believe that the ID in this case is the role's arn, am I correct?

Hi, I'm new to pulumi and have a question about refresh. Does refresh execute the program?

Guys, how do I get a resource that already exists? I am trying to get my AWS Cluster created in another project (therefore, another stack) so I am doing something like:

awsx.ecs.Cluster.getDefault({ id: '...', urn: '...'});

But I get the error UNKNOWN: missing required URN

Hi folks I've managed to get dependency tree file with

pulumi stack graph

. Now what kind of tools I can use to visualize this file?

Hi all! I'm configuring an autoscaling group for EC2 instances and I can't find the Termination protection parameter in the template parameters of awsx.autoscaling.AutoScalingGroup(templateParameters: {}). Where can I configure it?

What's the recommended way to deal with generating resource names? I create a bunch of servers, then security groups, then stuff their IPs into security group rules, so the group rule names are kind of three levels deep in terms of generating them. And they need to be unique, but they also need to be consistent between executions, or otherwise pulumi will be deleting old ones and creating new ones all the time. I cannot use machine names, because apparently I cannot refer to properties of other objects at the time when the names are generated. Any other ideas? (in case it matters: python, openstack)

Hey everyone! A question about AWS Elasticsearch service: AWS has some problems with removal of ES Domain, which happen randomly, and there's nothing we can do about it, it's a known issue they don't care about. I was wondering if there's a way to tell Pulumi to run the

delete

operation and NOT wait for it to finish? Because it can take hours for ES Domain to get deleted, and that's not acceptable, especially not in CI environment. Any ideas greatly appreciated, thanks!

[Resolved] Hi everyone - Has anyone tried setting up multiple projects (with stack references) when using an S3 backend? When we try to create multiple projects, their stacks show up bundled together on running

pulumi stack ls

, giving the impression that one S3 bucket only supports one project. If it is so, how can we do stack-references across projects? Also, we can't create stacks with the

org/program/stack

scheme that a fully-qualified stack name requires as per the documentation for stack references. This works fine when I use the Pulumi cloud backend. The larger question is whether org-project-stack data model and stack references are fully supported in non-Pulumi-cloud backends.

I think I will post more questions ( like this ) in Stack Overflow to increase Pulumi's tag popularity there. However, I'm trying a quick response so I will post it here as well: How can I remove imported resources from my stack without deleting them from aws?

I don't see any options in the documentation on how to delete imported resources from my stack. If I try to remove the resource's reference from my code I get the following error when running `pulumi up`:

error: Preview failed: refusing to delete protected resource 'urn:pulumi:dev::my-cloud-infrastructure::aws:iam/instanceProfile:InstanceProfile::EC2CodeDeploy'

Hi everyone - I have a dumb question: I have an azure AKS k8s cluster that was created through pulumi and I need to upgrade the kubernetes version (1.8.6 -> 1.8.14). Pulumi preview says it will update kubernetesVersion but it also says it will replace 2 other resources (pulumi😛roviders:kubernetes, kubernetes:core/v1:Namespace (my k8s namespace where all our development resources are deployed)) - is this a safe operation or am I going to lose anything? It also prints the following next to each "replace": [diff: ~kubeconfig], [diff: ~metadata,provider] (i'm super paranoid - if pulumi deletes this namespace i'm out of a job

[Resolved] Hi everyone. I have a question regarding moving the code from

index.ts

into separate files. Let's say I have a

postgres.ts

and

cluster.ts

. How would I go about telling pulumi to include those files? I've tried defining them in

tsconfig.json

without luck. I've got it working by exporting something from

postgres.ts

and then importing and exporting it again in

index.ts

. But this approach is more hassle than it's worth. Is there a correct way of doing this?

Hello guys Did anyone implemented/knows dynamic provider to work with istioctl directly?

Hello guys, is it possible to do a retry when an error is thrown when creating a resource. I have a problem that APIM try to get the swagger doc from the app service. It tries to get the swgger doc before swagger is up. I use deploy options with dependencies to the deploment oof the app service, when creating the APIM api

Has anyone noticed the removal of platform/language specific documentation? For example https://www.pulumi.com/docs/reference/pkg/dotnet/Pulumi.Gcp/ where users could see all type information for DotNet by Gcp API? The new documentation experience isn't so good. Thoughts?

Is there a way to skip deletion for a resource? For instance, instead of deleting appsettings of a web app, it just deletes the whole webapp?

What's the secret sauce in pulumi-aws repo that allows the projects in examples/ to use the current code in the repo, or does it always reach out to npm registry to download latest?

hello i have at the moment a issue with pulumi in gitlab-ci with gitlab-runner on aws .. Error: NoCredentialProviders: no valid providers in chain. Deprecated. - any idea ?? problem is also when we use default secrets-provider

I'm thinking about looking into more automation where I'm at, one part I would like to try out is creation of default setups when initiating a new project. What I have in mind is that based on some metadata a github repo should be created with some default data and access rights configured, along with that a default empty pulumi stack as well. Is that something where the pulumi automation api would make sense? Or would you use a dedicated repo with a "standard" pulumi project that did the creation of repositories and other github related stuff based on some meta data? Is it possible to create stacks in this case?

Hello again! Is there a safe way to interrupt and rollback (possibly destroying already made changes) a "pulumi up" deployment? I'm working with autoscaling groups and sometimes I make a mistake and want to stop the deployment process to fix it, but I know that if I just hit Ctrl+c I'll mess with the state file and corrupt my environment, but sometimes this deployment may take like 20 minutes!

I know that "pulumi cancel" tries to achieve this but the documentation is very clear that this may cause the same problems I had when hitting ctrl+c

@gorgeous-egg-16927 @calm-apple-46440 hi I found a k8s to yaml use case still failing after the fix was merged last year: https://github.com/pulumi/pulumi-kubernetes/issues/992

is there a way to have a temporary file as a resource? e.g. I'm doing dynamic resource that shells out commands and I need some cli options to be passed as path

I am getting following error while doing pulumi up import pulumi_azuread as azuread ModuleNotFoundError: No module named '*pulumi_azuread*' error: an unhandled error occurred: Program exited with non-zero exit code: 1 I tried installing pulumi_azuread module using below command. Still the error persist

pip install pulumi-azuread

having a problem with AWS SSO using v2 cli auth, and Pulumi, is it supported?

Guys, good evening! Rushing here again for help. I'm launching an auto scaling group and the underlying stack takes forever to be created. Jaxxtorm mentioned yesterday that I could set custom timeouts for resources, so I did it and this is my final autoscaling group instance definition:

const autoScalingGroup = new awsx.autoscaling.AutoScalingGroup(`${STACK}-myproject-baseline-asgroup`, {
        launchConfiguration: autoScalingLaunchConfiguration,
        templateParameters: {
            minSize: 1,
            healthCheckType: "EC2",
        },
        vpc: myVpc,
        subnetIds: myVpc.privateSubnetIds,
        targetGroups: [
            targetGroup
        ],
        onFailure: "DO_NOTHING"
    },
    {
        customTimeouts: {
            create: '5m',
            update: '5m',
        }
    }
);

However, this didn't work, and I'm still subject to the stack timeout. BUT, that is not the real issue. The real issue is that the stack never receives a success signal and, therefore, it always rolls back and gets deleted. This is the error message:

Received 0 SUCCESS signal(s) out of 1.  Unable to satisfy 100% MinSuccessfulInstancesPercent requirement

The stack is launched and an EC2 instance is spawned, I'm passing a user_data script to this EC2 instance that is working perfectly. After some research I found out about the cfn-signal script that can be used in CloudFormation stack templates, but I couldn't find any usage example of this use inside user_data scripts. Just to point out, I've also given all cloudformation permissions in the autoscaling launch configuration, to ensure that the stack can talk to the EC2 instance. Does anyone have a clue of whats going on?

Hi all, I’m getting an odd error that recently started occurring out of the blue.

pulumi:providers:random (default_3_3_0):
    error: no resource plugin 'random-v3.3.0' found in the workspace or on your $PATH, install the plugin using `pulumi plugin install resource random v3.3.0`

I’m using the python SDK with the requirements:

pulumi            2.20.0
pulumi-azure      3.45.0
pulumi-azuread    3.3.0
pulumi-kubernetes 2.8.0
pulumi-random     3.0.2

With the following plugins cached:

❯ pulumi plugin ls
NAME        KIND      VERSION  SIZE    INSTALLED    LAST USED
azure       resource  3.46.0   215 MB  6 days ago   6 days ago
azure       resource  3.45.0   214 MB  1 week ago   1 week ago
azure       resource  3.44.0   213 MB  2 weeks ago  2 weeks ago
azure       resource  3.42.0   211 MB  1 month ago  1 month ago
azure       resource  3.41.1   210 MB  1 month ago  1 month ago
azure       resource  3.3.0    175 MB  1 week ago   1 week ago
azuread     resource  3.3.0    43 MB   1 week ago   1 week ago
azuread     resource  3.2.0    43 MB   2 weeks ago  2 weeks ago
kubernetes  resource  2.8.1    70 MB   2 hours ago  2 hours ago
kubernetes  resource  2.8.0    70 MB   1 week ago   1 week ago
kubernetes  resource  2.7.7    70 MB   2 weeks ago  2 weeks ago
random      resource  3.0.2    31 MB   5 days ago   5 days ago
random      resource  3.0.1    31 MB   1 hour ago   1 hour ago

Any ideas? I’m not sure where it’s getting v3.3.0 for the random provider from!

Morning all, I'm working on a case study to implement pulumi over terraform. Does anyone have a spreadsheet/comparison of both tools?

Hello, I am evaluating Pulumi for Blob storage. Is it necessary to install Pulumi in the system or is there a way to directly use the libraries/APIs for Blob storage?