There's no way of importing resources yet @orange-airplane-70350 as far as I understood

Well, no easy way sorry, you need to manually tweak your checkpoint file by adding resources before being able to interact with them

Otherwise, you can still keep your existing resources and add new ones using pulumi but only the ones you'll have created with pulumi will be manageable that way

Can I not send a message to

aws.sqs.Queue

from JS?

Our latest blog post from @bitter-oil-46081 http://blog.pulumi.com/managing-github-webhooks-with-pulumi

Even if you aren't interested in GitHub Webhooks, I have some information on how you can use the dynamic provider to build your own custom resource provider with your application code. I use this to make it easy to register and delete the hook on a GitHub repository. There's actually a lot of cool stuff you can do there (we use it extensively in our testing).

Hi everyone, happy Friday. I just recently got started with Pulumi and I have a question about suitability for my project. I’m working on a TypeScript library that makes it easier to create and invoke serverless functions. The library is intended to be for scaling batch and stream processing, and less so for event handling. I’m wondering if Pulumi would be a suitable library to use internally for implementation.

After trying a bit of the tutorial I was struck by the need to install and use the pulumi command line tool, and login to the pulumi site for example. This wouldn’t be appropriate for users of my library - so can I use Pulumi in my library/app, but not require my users to in turn install pulumi?

@orange-airplane-70350 @able-tomato-34262 Great discussion on importing existing resources. Importing existing resource state is indeed a "work in progress" for us. We have a variety of GitHub issues tracking conversion tools more broadly, but I just created one to track the idea of a

pulumi import

command: https://github.com/pulumi/pulumi/issues/1635. If you want to do it by hand today, you can, but it's very much a manual process along the lines of what @able-tomato-34262 suggests (

pulumi stack export

your checkpoint state, add the empty resources to it,

pulumi stack import

it, and then run

pulumi refresh

to read in all the state). I realize this is cumbersome and if you'd like help, we're happy to spend some time with you.

pulumi import

would automate this process.

@big-piano-35669 How do you feel about adding functions for common policy type templates to

pulumi-aws

?

Is there any guidance on overriding the type of a field in the

resources.go

file of the AWS provider with an array type? I thought

tokens.NewArrayTypeToken(awsType(autoscalingMod+"/notificationType", "NotificationType"))

may have been it, but that panics at generation time complaining of a missing module delimiter. I don’t see anything of this form in either the GCP or AWS definitions.

This library has the potential to be useful: https://github.com/ip-num/ip-num

I’ve just published my first package of Pulumi code ported over from some of the Terraform examples I’ve published in the past. This one is a VPC component. There is some overlap with

pulumi-aws-infra

, but this has a few differences: - private/public address space is assigned in an assymetric fashion (as per “Practical VPC Design”), and subnets are automatically sized based on the available address space and number of AZs - a

PerAZ

option can be used for subnet distribution across all AZs in a region. No slack space is left since a second CIDR block can now be added. - VPC endpoints for S3 and DynamoDB can optionally be created - Flow logging can optionally be enabled - Tags can be applied to every resource - A private hosted zone can optionally be created and assigned via a DHCP Options Set It’s available here: https://www.npmjs.com/package/@operator-error/pulumi-aws-vpc (the org name is taken from our blog, https://operator-error.com), and on GitHub at: https://github.com/jen20/pulumi-aws-vpc. Any feedback is much appreciated!

Congrats and thanks - this looks great!

Incredible work! 💥

What's the current level of GCP support? Will I be able to create a simple Cloud Function with HTTP trigger? (and maybe pub/sub?)

Hey everyone, GitHub’s API is currently experiencing an outage which is causing https://app.pulumi.com to return erroneous 404s. We are looking into the situation.

Is it possible to adopt a less objectionable directory than

/opt

for

make install

on *nix? Something under ~ would be ideal. I’m generally not running with write access to

/opt

(it has its own ZFS dataset) and running the entire build elevated feels pretty nasty. It doesn’t particularly look like there’s a reason it needs to be machine global instead of per user?

hello! i'm new to pulumi. I think it's a great idea and was wondering if it's only meant for cloud resources or also for provisioning a server once it's spawned

Does anyone have an opinion on adding a

Tags

type defined like this: https://github.com/jen20/pulumi-aws-vpc/blob/master/src/index.ts#L13-L19 as an overlay for TypeScript? It would be helpful to have something common, as otherwise most modules end up redeclaring it. Something at the top level with ARN etc would make sense IMO

Hi all, I’ve spent the last couple of days diving in to Pulumi to get some AWS infrastructure spun up and have a React app deployed to it. It’s been a great experience so far and I’ve even managed to get it incorporated into my CircleCI workflow with relative ease. I do have a question, that I’ve been struggling with (and I briefly mentioned it to @big-piano-35669 on Twitter) regarding “immutable infrastructure”. This may be more a theoretical and conceptual misunderstanding on my behalf than something specifically Pulumi related, but I thought I’d ask since we’re obviously all here with similar goals. To my mind, one of the key tenets of immutable infrastructure is that when you do a re-deployment, you tend to spin up new infrastructure from scratch, deploy to that and then tear down the existing infra so that it’s replaced in it’s entirety. Pulumi seems to have a focus on updating existing infrastructure. E.g. if I change a property of an S3 bucket, preview/update appear to update the existing bucket in place, rather than create a new one. I also noticed this yesterday when I deleted something from the deployed infrastructure manually and then Pulumi couldn’t run the update. What am I missing/misunderstanding here? Would love to hear from anyone/everyone about this. Sorry for the long-winded first post 🙂

Hi all, I’ve just pushed (to GitHub, not to NPM) another new component for AWS: https://github.com/jen20/pulumi-aws-instance-profile - feedback appreciated! There’s a bit more to do in terms of making it accept things like a list of

Statements

optionally instead of complete `PolicyDocument`s, but this is a fairly common set of resources to find together in more traditional AWS infrastructure.

Hi, I know Pulimi corp was founded in 2017 but when did work on Pulumi started ? I swear I saw something about it somewhere but I can no longer find it 😞

The first commit according to Github is from 2017 as well I think @adventurous-jordan-10043

Yep I'm looking for something like first draft / POC / MVP of the whole concept behind Pulumi

If it's 2017 then it's 2017 but I thought it was a bit earlier than that

I believe the first record that I’m aware of it being “a thing” was this blog post, from June, 2017: http://joeduffyblog.com/2017/06/01/an-update-on-me-pulumi/ As far as a proof of concept or MVP of Pulumi, the product and technology has evolved a lot since last summer. It wasn’t until the Fall of last year that the programming model you see today really started to take shape. (Before then there were a lot of differences in how programming languages and the core libraries were handled. Those early versions were pretty rough to use…)

Latest from our blog: http://blog.pulumi.com/serving-a-static-website-on-aws-with-pulumi

Thanks to @colossal-beach-47527