Hello, quick question from newbie about projects and stacks naming. I have following folder structure (currently trying to migrate off terraform) that is kept in GIT.

└── infrastructure
    └── network
        ├── dev
        ├── prod
        └── stage
            ├── 01-host-project
            │   ├── Pulumi.stage.yaml
            │   └── Pulumi.yaml
            └── 02-host-vpc
                ├── Pulumi.stage.yaml
                └── Pulumi.yaml

So by default I am getting:

<organisation>/02-host-vpc/stage

as a stack name. What is the suggested approach to have project name aligned to my directory structure? Ideally something like:

<organisation>/infrastructure/network/02-host-vpc/stage

? Also I use this directory structure as a template for different projects/customers does it mean that for each separate deployment I need to have different

<organisation>

? Thanks!!😉

how do I get the string representation of a stack output? e.g. https://gist.github.com/ohlol/90cdf58e5a2ea49976470c06d8fb4afd

What is the best way to bootstrap Pulumi project via CLI in automated fashion (non-interactive)? I am running following script:

export PULUMI_CONFIG_PASSPHRASE=testtest

pulumi new gcp-typescript \
  --name 02-host-vpc \
  --description "GCP host VPC" \
  --stack stage \
  --secrets-provider passphrase \
  --yes \
  --non-interactive \
  --force

pulumi config set gcp:project "test"

This creates new project

Pulumi.yaml

and

Pulumi.stage.yaml

as required but it overwrites my

index.ts

file. Is there any way to stop that? Thanks 😉

In this tutorial https://www.pulumi.com/docs/tutorials/aws/aws-ts-ec2-provisioners/ what is the significance of the final two dashes after each cat line? Are they necessary or is that a typo?

$ cat rsa.pub | pulumi config set publicKey --
$ cat rsa | pulumi config set privateKey --secret --

is this a bug? I went to use some existing code, and now it's complaining about this code, saying it should be a string rather than bool, seems like a bug

networkInterfaces: [{
                    associatePublicIpAddress: false,
                    deleteOnTermination: true,

interface LaunchTemplateNetworkInterface {
        associatePublicIpAddress?: pulumi.Input<string>;
        deleteOnTermination?: pulumi.Input<boolean>;

am I doing something wrong with

ec2.get_subnet_ids()

in Python SDK? I pass

vpc_id=vpc.id

as I do with other resources but get an exception:

Exception: invoke of aws:ec2/getSubnetIds:getSubnetIds failed: "vpc_id": required field is not set ()

Hi. Quick question. I have this code:

const s3AccessKeyIdSecret = args.s3AccessKeyId || config.getSecret('s3AccessKeyId')
    const s3SecretAccessKeySecret = args.s3SecretAccessKey || config.getSecret('s3SecretAccessKey')

    const s3AccessKeyId = pulumi.all([s3AccessKeyIdSecret]).apply(([id]) => id)  
    const s3SecretAccessKey = pulumi.all([s3SecretAccessKeySecret]).apply(([secret]) => secret)

How can I simplify this? I just want to get the secrets from Pulumi secrets if I don't specify them in the class. I am new to Typescript. Also, I am using these secrets in two objects, but for some reason the variables are not "empty" only when accessed the first time for the first object. When accessed for the second object the variables yield no values - I can see this in the config map that is created without those keys only for the second object; the objects are defined in the same scope and the keys that use these secrets are the same. Why aren't the variables working properly when used the second time? Thanks!

Uhm is it just me or is Pulumi really slow?

Is it save to have a stack change from

gcp.projects.Services

to

gcp.projects.Service

? Pulumi shows it will delete the resource, but will that disable all of the APIs?

good afternoon, I have my code in typescript, everything turned out ok, but now I want to be able to run a script to copy files to my azure fileshare, only when I finish applying the changes. It can? How do I detect that I finish applying the changes, what is the callback? Is there an after apply method?

For the CI/CD pipline how do we check that the stack already exists? When I call

pulumi stack init my-stack

and the stack already exists pulumi throws an error

error: no stack named 'my-stack' found

I think pulumi cli should rather say something like

stack named 'my-stack' already exist. OK

hi

my pulumi on command line hangs.

and the login to "https://app.pulumi.com/" hangs too.

any problems on the server side?

Are there are any tutorials to create GCP cloud functions with Pulumi? 🤔

well the login now worked, but the console hangs, the previous call returned:

error: could not deserialize deployment: reading response from API: read tcp 192.168.0.119:52138->52.26.53.20:443: read: connection reset by peer
<http://github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).getTarget|github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).getTarget>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/httpstate/state.go:291
<http://github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).newUpdate|github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).newUpdate>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/httpstate/state.go:247
<http://github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).runEngineAction|github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).runEngineAction>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/httpstate/backend.go:951
<http://github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).apply|github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).apply>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/httpstate/backend.go:934
<http://github.com/pulumi/pulumi/pkg/backend.PreviewThenPrompt|github.com/pulumi/pulumi/pkg/backend.PreviewThenPrompt>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/apply.go:114
<http://github.com/pulumi/pulumi/pkg/backend.PreviewThenPromptThenExecute|github.com/pulumi/pulumi/pkg/backend.PreviewThenPromptThenExecute>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/apply.go:199
<http://github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).Refresh|github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudBackend).Refresh>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/httpstate/backend.go:819
<http://github.com/pulumi/pulumi/pkg/backend.RefreshStack|github.com/pulumi/pulumi/pkg/backend.RefreshStack>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/stack.go:87
<http://github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudStack).Refresh|github.com/pulumi/pulumi/pkg/backend/httpstate.(*cloudStack).Refresh>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/backend/httpstate/stack.go:152
<http://github.com/pulumi/pulumi/cmd.newRefreshCmd.func1|github.com/pulumi/pulumi/cmd.newRefreshCmd.func1>
/home/travis/gopath/src/github.com/pulumi/pulumi/cmd/refresh.go:127
<http://github.com/pulumi/pulumi/pkg/util/cmdutil.RunResultFunc.func1|github.com/pulumi/pulumi/pkg/util/cmdutil.RunResultFunc.func1>
/home/travis/gopath/src/github.com/pulumi/pulumi/pkg/util/cmdutil/exit.go:112
<http://github.com/spf13/cobra.(*Command).execute|github.com/spf13/cobra.(*Command).execute>
/home/travis/gopath/pkg/mod/github.com/spf13/cobra@v0.0.3/command.go:766
<http://github.com/spf13/cobra.(*Command).ExecuteC|github.com/spf13/cobra.(*Command).ExecuteC>
/home/travis/gopath/pkg/mod/github.com/spf13/cobra@v0.0.3/command.go:852
<http://github.com/spf13/cobra.(*Command).Execute|github.com/spf13/cobra.(*Command).Execute>
/home/travis/gopath/pkg/mod/github.com/spf13/cobra@v0.0.3/command.go:800
main.main
/home/travis/gopath/src/github.com/pulumi/pulumi/main.go:49
runtime.main
/home/travis/.gimme/versions/go1.13.9.linux.amd64/src/runtime/proc.go:203
runtime.goexit
/home/travis/.gimme/versions/go1.13.9.linux.amd64/src/runtime/asm_amd64.s:1357

Exist any event for execute method after apply changes (pulumi up - yes)?

Hi, I’m currently struggling with Pulumi and the cloudformation stack for my launch configuration (EC2 instances). Whenever I update the launch configuration (let’s say I update the user-data script) then it tries to update the cloud formation stack which timeout after 20 minutes (on the events section of the stack, I can see it’s stuck at Terminating instance(s) [i---------]; replacing with 1 new instance(s)), then it fails and rollback the cloudformation stack and thus I’m stuck with an unhealty stack. Edit : To be more specific, the “AWS::AutoScaling::AutoScalingGroup” ressource is stuck at “UPDATE_IN_PROGRESS” The only solution I have right now is to delete the cloudformation stack, pulumi refresh and then pulumi up which is OK because I’m still aat a development stage so I don’t mind if my EC2 instances are shutdown but it’s not the right solution when everything we’ll be in production. Anyone had this issue or an idea to resolve this? Thank

Obvisouly I just found out the root of the problem, well after some time spent in cloudtrail it appears the volume size of 8gb is smaller than my AMI snapshot (12gb) so I guess this is why it’s stuck. Maybe it can help someone in the future 🙂

Hi, I have a pulumi project that I pulled down from source control that was created by a teammate. It requires the AWS plugin, but I don't have that installed. It seems that it got installed for my teammate when he ran

pulumi new aws-python

. So now I want to install it via

pulumi plugin install resource aws VERSION

, but I cannot find any way to determine what the legal values of VERSION are... anyone able to point me in the right direction?

Hi folks, we’ll be hosting open office hours today @ 12pm PT in Zoom if anyone wants to join us! Bring Q/A on general Pulumi, AWS, GCP, Azure and Kubernetes, or stop by to say hello 👋 https://zoom.us/j/173438189?pwd=VmlFUFFRRmFvWHJsSTFmSjJKaDlRQT09

Pulumi is absolutely amazing 😄 I have written code to automate basically everything that my app needs to work in Kubernetes: cert manager, velero, nginx ingress, postgres clusters, pgadmin, redis, anycable, memcached, csi drivers. Now if I need to rebuild everything from scratch, I first create the Kubernetes cluster with Rancher, and then I run a single

pulumi update

command. Finally, I deploy my app with CI/CD. Super awesome! I am doing some final tests of my code and then I will rebuild my actual cluster this way so I can maintain it with Pulumi in the future

Office hours have started if folks would like to join! https://zoom.us/j/173438189?pwd=VmlFUFFRRmFvWHJsSTFmSjJKaDlRQT09

I have two stacks, test and prod, and I have two files called Pulumi.test.yaml and Pulumi.prod.yaml with the different config and secrets. However I have one index.ts only with what I want Pulumi to deploy. Can I have two different files instead of the single index.ts and specify which one to use for each stack?

Is the current best practice for secrets inside magic/serialized functions to use a SSM Parameter Store / Secrets Manager resource and include that in the function via SDK? Is there anyway to ensure that the secret is serialized outside of the handler so it stays up inside Lambda execution context? Is there a simpler way or a component in the works for handling the above?

got a quick question: does k8s pulumi support "kubectl user impersonation", so something like

kubectl --as=system:serviceaccount:ns-sourcegraph:fake-user -n ns-sourcegraph apply -k .

where fake-user is a service account with some policies attached.