👋 general question about plugins. I manually install the versions of the plugins that i want, by doing:

pulumi plugin install resource aws v3.2.0
pulumi plugin install resource kubernetes v2.4.3

But sometimes when I then run

pulumi up

, pulumi starts installing different versions of the same plugins..

pulumi up
Previewing update (dev):
[resource plugin aws-3.2.1] installing
Downloading plugin: 0 B / 68.86 MiB [----------------------------------]   0.00%
[resource plugin kubernetes-2.5.0] installing

Can I disable that? Where is it configured?

if I have a multi-file node project (multiple local modules), in a Lambda setting, what's the proper way to have Pulumi deploy them along with the definition of the Lambda callback?

Hey guys, I think I have hit an infinite recursion problem, activating the service usage API fails with pulumi because… well because the service usage API has not been activated yet!

gcp:projects:Service (EnableServiceUsage):
Step #9 - "Pulumi Create Stack":     error: Error when reading or editing Project Service : Request "List Project Services my-project" returned error: Failed to list enabled services for project my-project: googleapi: Error 403: Service Usage API has not been used in project XYZ before or it is disabled. Enable it by visiting XYZ then retry. If you enabled this API recently, wait a few minutes for the action to propagate to our systems and retry., accessNotConfigured

sup guys trying to launch test example of aws sns on empty project with the latest libraries, has panic accessing nil (golang): https://gist.github.com/afoninsky/250c1dbb96ff00ceaa2d830f0535fc93 I wonder if somebody met similar behaviour?

I'm running into an issue with creating a Vault database connection due to errors with parsing the connection URL. I'm not sure if it's because I'm trying to pass in the host_url as an Output object and it's not resolving before trying to pass it to Vault?

db_config.db_connection = 'postgres://{{db_user}}:{{db_pass}}@{{db_host}}:{{db_port}}/{{db_name}}'        
self.db_connection = database.SecretBackendConnection(
            f'{db_config.db_name}-database-connection',
            opts=resource_opts.merge(ResourceOptions(parent=self.db_mount)),
            backend=db_config.mount_point,
            verify_connection=db_config.verify_connection,
            name=db_config.db_name,
            data={
                'username': db_config.db_admin_username,
                'password': db_config.db_admin_password,
                'db_host': db_config.db_host,
                'db_port': db_config.db_port,
                'db_name': db_config.db_name
            },
            **{db_config.db_type: {
                'connection_url': db_config.db_connection
            }}
        )

The host_url is the

address

attribute of a created RDS instance.

say you have a large pulumi project and something goes bananas, it happens not very often but sometimes. You need to export the stack, remove pending operations but then you may remove operations that other resources are dependant on and you end in a loop that takes hours to solve. Is there are faster way? can I just remove manually all resources and then tell pulumi to start from clean slate?

There's apparently something called

lifecycle

which for example could let me specify "*don't* delete this VM when

user_data

changes", but the only mentions of it I found are in https://www.pulumi.com/docs/reference/pkg/aws/ec2/launchconfiguration/ which points to https://www.terraform.io/docs/configuration/resources.html. So how does one specify this? (in python, if that makes a difference)

Hi everyone. I tried to change my stack passphrase and now I'm getting the following error:  

Preview failed: unable to discover AWS AccessKeyID and/or SecretAccessKey

I'm not being able to fix this issue, I reviewed my .aws/credentials and .pulumi/credentials but no idea. Any idea?

How does Pulumi detect changes with FileArchive/AWS Lambda function resoruces? Every time I do a deployment in a CI/CD pipeline all lambdas redeploy regardless of if the actual content has changed. Looking to find a way around this.

I am deploying docker container on our docker host which support docker API version 1.39 but not latest one. From docker cli, we can set version and can connect to docker host.

export DOCKER_API_VERSION=1.39

However with pulumi, its not working There is no option to set API version https://www.pulumi.com/docs/reference/pkg/nodejs/pulumi/docker/config/ When deploy with pulumi I am getting error

error: Error pinging Docker server: request returned Internal Server Error for API route and version <https://DNS>:PORT/_ping, check if the server supports the requested API version`

I am using below config with pulumi

docker:host: "<tcp://DNS:PORT>
docker:certPath: "valid path to docker-cert"
docker:tlsVerify: 1
docker:apiVersion: 1.39

Any help here?

our docker host do not support latest version 1.40 so how can I set lower API version in Pulumi

➜  mta git:(pulumi) ✗ export DOCKER_API_VERSION=1.40
➜  mta git:(pulumi) ✗ docker ps
Error response from daemon: client version 1.40 is too new. Maximum supported API version is 1.39
➜  mta git:(pulumi) ✗

Hey, I have a scenario of a production environment with many single tenant EC2 based copies of our application. Each instance is a different tenant. Couple of questions: • Design-wise, would I create a stack for each one or one big stack for all? • If the first is prefered, what is the best way to create and update multiple stacks at once?

Is there a full example of a HTTP API Lambda integration using

aws.apigatewayv2

? I’ve got a pretty simple

awsx.apigateway

with only Lambda proxy routes I’d like to migrate to v2

Q: I have just transferred a project/stack from my personal account to an org account using the Transfer ownership button in the web ui. For example:

my_personal_account/project_name/prod

to

my_org/project_name/prod

Now I want to switch my local code to the new org stack:

pulumi stack select my_org/project_name/prod
# => error: no stack named 'my_org/prod' found

Is this some kind of bug? Or I have missed any important step here. Many thanks in advance

Q I am creating a temporary password for AWS IAM user. I follow the instruction by adding `UserLoginProfile`:

let user = new aws.iam.User("totoro", {
  name: "totoro",
});

// force new users to reset their password for the initial sign in
new aws.iam.UserLoginProfile(`login-profile-for-totoro`, {
  user: user.name,
  passwordResetRequired: true,
  pgpKey: `WHY_DO_I_HAVE_TO_PROVIDE_THIS?`
}, {
  parent: user,
  dependsOn: [user]
})

I am wondering why this resource requires me to use

pgpKey

to encrypt/decrypt password (/with Keybase.io) when there is already KMS mechanism provided out of the box by Pulumi. Many thanks in advance Reported in https://github.com/pulumi/pulumi-aws/issues/1106

Has anyone figured out how to deploy a helm chart with CRDs to multiple clusters in the same pulumi stack? (https://github.com/pulumi/pulumi-kubernetes/issues/1225#issuecomment-684237439)

Does anybody know how to use pulumi.all() with more than 8 parameters please?

Is the config, as loaded by Pulumi in the app, available during a unit test using the mock functions set up in a Typescript test project? I'm trying to write tests for my Pulumi app, but my

config.require()

calls are throwing an error even though I've passed the project name (and stack name, for that matter) to

pulumi.runtime.setMocks()

.

I got a project I haven't touched in a while, it contains a cloudformation stack I'd like to update. It detects it needs updating due to the template being different. When I do

pulumi up -t urn:to:stack

it looks like it performs the update but doesn't actually do anything. Has anyone dealt with this before?

I wonder if you guys can share some experiences when you end with a broken stack with no other solution than remove everything and start all over?. It happens me quite often, it can be some resource that did not complete properly, or something weird that makes the stack end in a state where you basically need to export-> correct problem -> import again. However in a big stack with many dependencies it is just virtually impossible to correct the problem manually. This leaves no other option than remove the stack, remove all resources manually and start over. What bothers me is, how you deal with this situation if it happens on production?

Has anyone encountered the issue that pulumi shows it has created a resource (in this case, an AWS CloudWatch Log Group), but no resource is indicated in the AWS Console?

Hi everyone. I'm trying to create eks and deploy couple of helm_v3 charts on top of it in one stack. Stack is written on python, I followed this example - https://github.com/pulumi/examples/tree/master/aws-py-eks. The question is, how can I export kubeconfig with

python

sdk? I see that

ts

has

generateKubeconfig

- https://github.com/pulumi/pulumi-eks/blob/master/nodejs/eks/cluster.ts#L182. But nothing for python. Any ideas?

is there a way to use a terraform provider with pulumi that isn't listed in https://www.pulumi.com/docs/reference/pkg/ ?

does the pulumi-terraform-bridge basically compile pulumi modules from terraform plugins? that part is kinda unclear to me.

hi, we've recently run into an issue with setting up Azure Service Bus queues in the premium tier/SKU the Pulumi docs says that EnablePartitioning MUST be set to true for Premium, but this contradicts what Microsoft says stating that Premium does NOT support it at all. So despite following the Pulumi docs, Azure forces it to false and each update recreates the queues because it trying to set it to true and gets ignored. https://www.pulumi.com/docs/reference/pkg/azure/servicebus/queue/#state_enablepartitioning_csharp The above docs where it says it must be set https://docs.microsoft.com/en-us/azure/service-bus-messaging/service-bus-partitioning#partitioned-entities-limitations Microsoft stating it's not supported on premium

for adding tags to resources, per this article, should this work with custom resources (extending pulumi.ComponentResource)? e.g. a resource which creates an @pulumi/azure/storage/Account should then have auto-tags added to the account on creation?

Hi. I believe it was asked already but I cannot find it anywhere in the pulumi docs or by searching google. Does Pulumi support applying changes from a previously generated preview? (like in Terraform where the plan can be stored in a file and then terraform apply command applies what was stored in the file)

We can pass env variable to kubernetes deployment using

new k8s.apps.v1.Deployment{
....
 containers: [
 {
    env: [
    {
       name: "db_host",
       value: config.get("db_host")
     }
  }
}

Similarly I need to pass env variable to docker container.

new docker.Container("container_name", {
    .....
    envs: [

    ]
}

But

env

in

docker.Container

is list of string and not key /value pair. Any idea how can I pass env variable with Container?

Is there any parameters to finish pulumi script fast after error? it takes 10 mins to complete otherwise..