How do people manage state when they use blob storage like GCS?

Looks like I have to create a bucket per service if I want fine-grained access control. Wondered if anyone had any insight to usage patterns.

At some point, I’d like to write a Kubernetes secret backend storage akin to how Helm does it

Question about organizing Pulumi projects & stacks. If we wanted to build out our infra using a simpler project/stack organization at first (1 project, and 1 stack for each environment, Dev/Staging/Prod), would we be able to migrate to a more modular setup at a later time? It would be nice to separate resources that are expected to change often from the resources that are not, but we don't necessarily want to take on that extra complexity up front.

Also, I’ve created the following for (hopefully) a bit of discussion: https://github.com/pulumi/pulumi/issues/6667

(About having a ‘stateless mode’)

This message was deleted.

Morning, how can I change the version of node that pulumi tells lambda to use from TypeScript?

it's the runtime arg in the lambda

Thanks Paul.

👌

(sorry, I was sleeping - late night working!)

Hi guys! I have a problem to use helm chart. I can't configure secret password in secret created. every time I deploy it it changes the password, I can't pass a password to the helm chart. I saw that there are some discussions on the topic, is this really a limitation today? Thank you

Have you got an example @narrow-airport-65967?

When trying to run

pulumi up

I am receiving the error :

error: could not load plugin for aws provider 'urn:pulumi:dev::permissions::pulumi:providers:aws::default': no resource plugin 'aws-v3.31.0' found in the workspace or on your $PATH, install the plugin using `pulumi plugin install resource aws v3.31.0`

But when trying to install the plugin I receive another error:

error: [resource plugin aws-3.31.0] downloading from : 403 HTTP error fetching plugin from <https://get.pulumi.com/releases/plugins/pulumi-resource-aws-v3.31.0-darwin-arm64.tar.gz>

As you might notice, I am on a MacBook Pro with the new M1 chip (hence, arm). I guess specifically this version of the plugin has not been compiled for this architecture, since I am able to install other versions of the plugin without any issue (e.g.

3.35.0

,

3.34.2

, and

3.34.0

) Any suggestions of how to handle this?

any reason why the stack names are arbitrarily limited to 100 characters? https://github.com/pulumi/pulumi/blob/master/pkg/backend/filestate/backend.go#L265

i know 100 may seem like a lot, but in my case with microstacks i try to have the project, account, company, region, cluster, env and variant name all in the stack name, I just hit 106 and it broke 😞 - seems like 255 would be a nice number to use instead, or whatever 255 minus

Pulumi

and

.yaml

would be ~244

Anyone getting errors from azure for no reason lately? Im getting an error access denied but Im accessing the content Im intending to change...lol weird. 🤔

Hi! So I'm trying to deploy the static content for my website using pulumi. I'm a bit confused though. After I build my app, I create BucketObjects for every file in the

dist

folder and that uploads them to the bucket just fine. However the next time i build my app, that files in the dist folder are different so pulumi tries to delete all the old files and upload the new ones (because the old files are no longer in the dist folder). Is there a way to keep the old files in dist around? It will break my app during deploy if the old static content is deleted before the new version of my server is deployed (its an SSR'd SPA). Thanks!

Hi! I'm working on a component resource (I'm still learning) and I'm wondering what the 'remote' flag means for ComponentResource (https://www.pulumi.com/docs/reference/pkg/nodejs/pulumi/pulumi/#ComponentResource-constructor)? (I also see this constructor param for the Resource class). How should I decide if it needs to be true or false?

Hello everyone, I am a beginner in Pulumi and I am testing https://github.com/pulumi/examples/tree/master/gcp-py-network-component I have configured everything well, except that when I do my pulumi up it gives me this : error: Configuration 'gcp-py-network-component:subnet_cidr_blocks' value '172.2.0.0/16' is not a valid 'JSON object' error: an unhandled error occurred: Program exited with non-zero exit code: 1 Do you know the exact syntax to declare a cidr in the pulumi config file? I typed this command : pulumi config set gcp-py-network-component:subnet_cidr_blocks '172.2.0.0/16' And here is the return in my Pulumi.dev.yaml file :

config:
gcp-py-network-component:subnet_cidr_blocks: '172.2.0.0/16'

Would you have an idea ? Thanks in advance for your answer

Is it possible to migrate a stack from one project to another? Would it just be

stack export

and

stack import

? Specifically I am using the S3 backend with namespaced stacks such as

<http://infastructure.operations.apps.QA|infastructure.operations.apps.QA>

and

<http://infrastructure.consul.apps.QA|infrastructure.consul.apps.QA>

I want to move the resources that are managed by the

operations

project to be owned by the

consul

project.

If I have to I can just destroy and rebuild, but that would be a bigger lift than just migrating the state.

DNS outage in Azure

This seems to be causing my code issues authenticating against storage blob which isnt allowing me to update my stack

Is it possible to get the name of the organization that owns the stack programmatically? Something a la

pulumi.get_stack()

. I’m not seeing anything in the docs.

@tall-librarian-49374 I noticed that pulumi stores stack history files in blob storage. Is it possible to somehow have pulumi restore infra state with those files? like lets say something fails to builds and you need to roll back.

Anyone ever see something like

tderr: error: could not load plugin for aws provider 'urn:pulumi:development::nacelle-services-dilithium::pulumi:providers:aws::default_3_35_0_alpha_1617104857'

? We aren’t using the alpha provider anymore and have updated the version in package.json (and package-lock.json)