Im pretty happy with provisioning DB server and databases with Pulumi. But I was wondering if I can (or should) also create the DB users that are to use said databases (and configured in backend applications). Or do you go back to plain SQL for that?

hey there, is there a way to force pulumi to delete rds/elasticache clusters on a destroy regardless of whether stuff is backed up or not

I'm trying to use the Vault provider and create a Policy object. It looks like it currently only supports an HCL formatted string, despite Vault being able to handle the equivalent JSON. I found this issue (https://github.com/pulumi/pulumi-vault/issues/10) asking about supporting an object (or even a JSON formatted string). Does anyone have pointers on steps to add that capability?

hey guys, I'm having issues tearing down a vpc, the internet gateway times out and i think it has something to do with dependency chains re Elastic ips and routetables. Is there any example anywhere as to how it should be chained together

Pulumi throwing an error when updating one of my kubernetes stacks: error: could not load plugin for kubernetes provider 'urnpulumistage:pod autoscalerpulumiproviderskubernetes:default_2_5_0': failed to load plugin /home/shaw/.pulumi/plugins/resource-kubernetes-v2.5.0/pulumi-resource-kubernetes: plugin not found  master  pod-autoscaler  pulumi plugin ls NAME KIND VERSION SIZE INSTALLED LAST USED aws resource 3.16.0 252 MB n/a 8 minutes ago aws resource 3.1.0 235 MB n/a 27 minutes ago docker resource 2.3.0 46 MB n/a 27 minutes ago kubernetes resource 2.7.2 79 MB n/a 7 minutes ago kubernetes resource 2.6.2 78 MB n/a 27 minutes ago kubernetes resource 2.5.0 66 MB n/a 24 minutes ago TOTAL plugin cache size: 756 MB any ideas?

running

pulumi up

gives me this error after ~10 minutes:

error waiting for service (service arn) to reach a steady state: ResourceNotReady: exceeded wait attempts

how can I debug why its taking so long? the Fargate service does seem available via the public endpoint so not sure if this is waiting for the right signal.

Why is an

apply

callback called during preview when the value is not available? Is this broken functionality?

To access the raw value of an output, and transform that value into a new value, use apply . This method accepts a callback that will be eventually invoked with the raw value, once it is available.

Why can't I do the following in my Github action?

- uses: <docker://pulumi/actions>
  with:
    args: preview --stack Safira/sandbox

It tells me

No stack configured for branch 'refs/pull/698/merge'

but I don't want that stack associated with a certain branch, I'm running it on a trigger word in pull requests.

Hello everyone, I want to get some feedback about performance from other pulumi users. I've already split my project into several Pulumi projects, but I've come to a point now where it would be very time consuming for me to divide it up further. For example, one of my Pulumi projects has 206 resources, mostly Kubernetes resources, some ComponentResources, and some dynamic resource providers. A 'pulumi up' takes 7-8 minutes even if nothing has changed. (Interestingly, when the command is finished, a "Duration" is output which is only half of that.) Opening the Pulumi web console overview page of that stack takes 20 seconds and switching to resource page takes another 20 seconds. Is that what I should expect from projects of this size? (Performance command line/web console) What experiences do you have with the project size and the resulting performance? What would you say is the max. number of resources per project that can be managed fluently? If you split your resources across many smaller projects, how do you handle dependencies and updates that affect other projects? BTW: I'm using typescript here, but I think at least the performance of the web console doesn't depend on the implementation language.

Hi, I'm trying to create an Azure web app for containers using the following code:

this.linuxAppServicePlan = new azure_nextgen.web.latest.AppServicePlan(
      name,
      {
        name: name,
        kind: "linux",
        location: this.resourceGroup.location,
        resourceGroupName: this.resourceGroup.name,
        sku: {
          name: "B1",
          tier: "Basic",
          size: "B1",
        },
      }
    );

    this.webApp = new azure_nextgen.web.latest.WebApp(name, {
      name: name,
      kind: "app,linux,container",
      serverFarmId: this.linuxAppServicePlan.id,
      location: this.resourceGroup.location,
      resourceGroupName: this.resourceGroup.name,
      httpsOnly: true,
    });

I created these resources in the portal first to take a look to the ARM template. The key element seems to be the

kind

as

linux

for the service plan and

app,linux,container

for the web app. However, Pulumi seems to be creating a windows app plan and a standard web app. Do you have an idea of what could be wrong?

Hey guys, I wanted to check whether this is a valid use case for Pulumi or not (and if you have any tips on how to go about it) .. In my company, we have lots of projects being tested and we'd like to be able to spin up a new test environment per project and then just tear it down once it's done.. the environments would generally be the same (several databases, a couple of app servers, a redis instance, an SSL certificate and hostname plus a load balancer). I'm in the process of setting up my first pulumi 'stack' but i'm wondering what the next step, once that's done, to be able to 'reuse' that stack for each new test environment that we need to stand up?

@ambitious-father-68746 watching your video from DOXLON 😄

nice comparison between the tools!

Thank you! Hopefully you didn't find it too boring, I lost track of time and went 10 minutes further than I should've.

Hey all, best person to reach out to about Octopus Deploy Step Template?

hi all!

i'm facing a problem on gcp

+  gcp:storage:BucketIAMMember role-legacy creating error: Error applying IAM policy for storage bucket "b/sky-it-backup-mysql-prod": Error setting IAM policy for storage bucket "b/sky-it-backup-mysql-prod": googleapi: Error 412: Precondition Failed, conditionNotMet
 +  gcp:storage:BucketIAMMember role-legacy **creating failed** error: Error applying IAM policy for storage bucket "b/sky-it-backup-mysql-prod": Error setting IAM policy for storage bucket "b/sky-it-backup-mysql-prod": googleapi: Error 412: Precondition Failed, conditionNotMet

what conditionNotMet mean?

Is there a way to know if we're running

pulumi preview

or

pulumi update

?

Hey, I just try the crd2pulumi tool. It’s really cool. I just doesn’t understand this

utilities.getVersion();

Just wanna understand why this is done inside the constructor of the resources. We move the generated files into our own repository and we are getting an error about this

i’m seeing this error when running `pulumi up -y`:

Error initializing PostgreSQL client: error detecting capabilities: error PostgreSQL version: EOF

seems similar to this issue: https://github.com/pulumi/pulumi/issues/3166 Any idea on why this might be happening?

have a question about pulumi pricing. is it possible to use it for free at scale? i.e. with lots of developers? say if you use s3 buckets for state? kinda like the "free" tf?

Hey folks! 👋😄 I created a dynamic provider and it works very well, except that whenever I change the implementation of the provider even slightly (a change that has absolutely no impact on outputs), pulumi updates the provider associated with all resources created with it, as if it was a totally different provider (maybe that’s the way dynamic providers are identified, by a hash of their code or something? so if the code changes slightly, it’s considered a totally different provider?) And that’s a problem because afterwards all previous state is no longer recognized (it’s undefined), so it always tries to update the resources at every run. Is there maybe a way to identify the dynamic provider/resource with a fixed known type (like we can do with

CustomComponent

resources when we pass a custom type string to the base constructor)? Maybe that will help the state to smoothly transition from older to newer version of same provider?

Question about API Gateway. Is it possible to provision an API Gateway with one repo, and setup the routes and lambdas for it in another repo using StackReferences?

Hey, who can answer question about

crd2pulumi

?

Hey all! I’m new here and still learning the ropes. Probably a silly question: do stacks need to have globally unique names and not just unique within a project?

Help! I originally signed up for app.pulumi.com using GitHub, but I need to transition that account to username/email. I reached out to support but haven't heard back. Is there anyone here who can help?

Hello fellow Pulumi-ers 😎. I did not see a "Jobs" channel in this slack, so I hope you will excuse me posting this here. The ACCURE team and I have been using Pulumi heavily for a few months now and we are really appreciating it! Now we're expanding our infrastructure team and are planning to bring more developers into the world of Pulumi. So if any of you are looking to get involved in a early-phase start-up in the heart of Europe (or know someone who is), then please free to get in touch.

Hey! Can I define specific IAM permissions for lambdas deployed in an API. gateway somehow?

I'd like to have multiple "projects", e.g. "github", "aws vpc", "aws eks cluster" in a single git repo that i can

up

independently from each other (or all together if i choose to). is there a recommended way on how i should organize this? so far i'm thinking of having a bunch of

Pulumi.project_stack.yaml

files and having the

index

file choose which codepaths to execute based on the config properties. but this seems not quite ideal I would be fine having everything in a single "project" but I'm concerned about long

preview

times if it needs to grab the state of every resource. or has this not been a big issue?

has anyone tried to use Pulumi to automate baremetal deployments?