I hope you guys are well. Actually I am doing a comparison by comparing features of pulumi with aws cdk. I am finding it difficult to compile up all the extra features which pulumi gives over aws cdk. Can anyone please help me out by highlighting some of the features?

why should I prefer pulumi for my projects, plus what are the major differences between paid and non paid version

How are project groups defined? I'd apparently created a project group once but can't seem to create a group of projects a second time? Not sure how.

I'm sure there was a readme on using parameters to populate a pulumi spec, but can't seem to find it now. Anyone got a link?

is there a specific channel to talk about the

pulumi/actions

docker container?

I'm having problems understanding setting up an application load balancer's AccessLogs in regards to the bucket, is this documentation just out of date? It seems like it's looking for an Input<string>, won't accept an S3.Bucket. Is it just looking for the bucket's name, or the ID? https://www.pulumi.com/docs/reference/pkg/dotnet/Pulumi.Aws/Pulumi.Aws.ApplicationLoadBalancing.LoadBalancer.html

AccessLogs = new Aws.LB.Inputs.LoadBalancerAccessLogsArgs
        {
            Bucket = aws_s3_bucket.Lb_logs.Bucket,
            Enabled = true,
            Prefix = "test-lb",
        },

When creating t2.micro instances I get a complaint sometimes that the size is not available in an availability zone, and to choose a different one or don't set one. I have not set any AZ explicitly (in older versions of my script--I "solved" the issue by forcing one). It seems like pulumi is distributing across AZs by default in AWS--is there a way to turn off that behaviour?

Has anyone tried out doing conditional adoption of existing resources? For instance, can you do something like:

existing_vpc = ec2.get_vpc(tags={'Name': 'my-vpc-name'})
if existing_vpc and existing_vpc.not_in_stack():
    import_opts = ResourceOption(import_=existing_vpc.id)

Where the

not_in_stack

functionality can do some broad lookup of whether Pulumi is already managing the given VPC?

So something interesting happened just now. I was trying to manage a secret using pulumi as 

pulumi config set --path "path/of/secret-key/"

 and entered a weird looking 16 char 

pwgen

 random password. And Pulumi detected that it looks like a password and asked me to use 

--secret

 flag to save it securely. How cool is that! 🙂 Pleasant surprise. And a DX.

So what’s the proper way (if any) of inspecting an Output type in Python? Printing it doesn’t give a pythonic string and I’ve tried numerous variations of apply() with no success either. Makes troubleshooting really difficult.

Hey there. Has anyone found a good pattern for a self refreshing “end date” on azure service principals? I don’t want to hardcode an expiry date for the service principal. Is there a way to do a

Date.now + N days

that won’t update on every deployment, perhaps only updating when it’s within a week of the expiry date?

Hi guys. I'm just getting started with pulumi but am facing an issue.

I'm creating an ecs cluster

ecs_cluster = aws.ecs.Cluster("ecsCluster", name="cluster-name", tags=default_tags)

and trying to refer to it using

ecs_cluster.name

(tried

ecs_cluster.name.apply(lambda name: name)

too.

I'm getting that as

<pulumi.output.Output object at 0x7f717c021d30>

I guess I'm facing the same issue as @mysterious-pilot-36551

Hi all. I have a variable for an API token defined like this:

const apiToken = pulumi.output(args.apiToken || config.getSecret('apiToken') || "blah")

then I am trying to interpolate its value in a string:

secretInline: `HCLOUD_API_TOKEN: '${apiToken}'`

Instead of the expected value, the following is stored (it's a Kubernetes secret):

Calling [toString] on an [Output<T>] is not supported.
To get the value of an Output<T> as an Output<string> consider either: 1: o.apply(v => `prefix${v}suffix`) 2: pulumi.interpolate `prefix${v}suffix`
See <https://pulumi.io/help/outputs> for more details. This function may throw in a future version of @pulumi/pulumi.

How can I fix? I've read a bit and tried adding 

.apply(v => v)

 but nothing changes. Thanks!

Hi guys, i'm new to using Pulumi on the pull down step

pulumi new aws-python -s tpilvelis-gw/aws-python/dev

I'm getting error

error: unable to open bucket <azblob://pulumi>: azureblob.OpenBucket: accountName is required

is this on the pulumi side?

Anyone? I still haven't been able to resolve it. I'm probably missing something very basic...

Help guys 🙂, yesterday I deployed a stack when logged in locally, and everything was successful, but today, when I go to the directory and

pulumi stack ls

, it doesn't show my stack (it's present in

~/.pulumi/stacks

and also my

Pulum.<stack>.json

is present in my directory); what's happening? --- UPDATE --- It's confusing for me as to why it should matter, but if I set

PULUMI_CONFIG_PASSPHRASE

, my stack is displayed

Is there a flag to turn off the random string suffix that gets appended to objects pulumi creates? I have a few persistent S3 buckets that I don’t want the random strings on but the rest of my items should have it.

A tip regarding slow performance in Pulumi programs that use the non-default Kubernetes provider: Pulumi might also be trying to connect to the ambient cluster from kubeconfig (and timing out if it points to a non-existent cluster). I found that the performance of my Pulumi program increased significantly by explicitly setting the ambient context to a dummy value.

config:  
  kubernetes:context: none

The odd thing is that all my resources from

k8s

have an explicit provider set, and so the ambient provider shouldn’t be needed. FYI.

is it possible to ignore tag keys that are unknown at compile time? Eg. I want to ignore EKS cluster ownership tags on AWS VPC subnets, but they contain the cluster name which is dynamic. How can I filter those tags out so they don’t get deleted? Able to filter the keys on prefix (

<http://kubernetes.io/cluster/|kubernetes.io/cluster/>

) would be ideal, but because they’re on the resource as an output and not in the

props

I can’t seem to use stack transformations.

is there a way of previewing a change, getting a snapshot and then applying that snapshot. i know theres

pulumi preview -j > plan.json

but i cant find the equivalent of

terraform apply plan

👋 I’m here! What’d I miss?

I found this pearl but cannot get it to compile -> https://www.pulumi.com/blog/program-kubernetes-with-11-cloud-native-pulumi-pearls/#3-inject-envoy-sidecars-using-abstraction

src/microservice/rex-common/base/BaseMicroservice.ts(5,21): error TS2532: Object is possibly 'undefined'.

src/microservice/rex-common/base/BaseMicroservice.ts(5,31): error TS2339: Property 'template' does not exist on type 'Input<DeploymentSpec>'.

The errors refer to the following line:

const pod = args.spec.template.spec;

Any ideas why I started getting this? (Typescript)

pulumi:pulumi:Stack (etl-test4):
    warning: resource plugin azure is expected to have version >=3.9.2, but has ; the wrong version may be on your path, or this may be a bug in the plugin
    warning: resource plugin azure is expected to have version >=3.9.2, but has ; the wrong version may be on your path, or this may be a bug in the plugin

I do pulumi stack export and then pulumi stack import immediately after, not modifying the file and get following error: "invalid character 'ÿ' looking for beginning of value" on Windows (powershell terminal) Is there a way to control/fix the encoding?

@pulumi/kubernetes

has issue to destroy the stack for “operator” style deployments. Probably it doesn’t know about pods/services that operator makes, so pulumi always fails to destroy the k8s. Any ideas, workarounds?