In a multi tenant system, where each customer has one instance of our application running?

There are several issues on the pulumi GitHub about pulumi not working when windows is in another language than English because of special characters. The only workaround is changing the OS language. I don't know since when this stop working but it was working in pulumi v1. I am organizing a workshop about pulumi in my company and tell the participants to all change their OS language doesn't seem a very good option but I don't see another one. any idea when this will be corrected ?

Got a deployment that appears to upload a file to my S3 bucket, and the file is created, but it's empty. The file has data on disk but in S3. Pulumi doesn't report any errors. Any ideas besides comb through my cloudtrail logs?

Would folks here recommend Pulumi to handle deploying apps to Kubernetes? I'm using Helm v2 right now and there's some trouble with upgrading to k8s v1.16. I'm not a fan of Helm in general because of the server-side components like Tiller and configmaps in

kube-system

, plus Go templating is horrendous (causing me cross my fingers). I'd rather not try upgrading to Helm v3 as that's likely to break the world and require backups and restore considerations. I'm wondering if Pulumi is a viable alternative to Helm v3, jsonett, ksonett, YAML/JSON with low-brow template system (e.g. envsubst), dhall-kubernetes, TF, etc.

Is it possible to get a GitHub Actions build artifact? Doesn't look like the GitHub package has that currently available? (cc @broad-dog-22463 I think you worked on that package?)

hi all. I would like to setup custom provider fields like in terraform:

I don’t find the way in pulumi.

may you drive me to an example or documentation?

I know this should be straight forward, but I’m struggling, how do I attach a IAM Role to an ec2 instance? I know I attach an instance profile, but what if I want to use a managed policy?

hi all trying to implement additional logic of building and pushing docker images, have some issues with implementation of the following: - trigger build/push only if such tag/hash does not exists locally - trigger build/push only if such tag/hash does not exist in remote hub any advices how to do it in Pulumi exactly? thank you

is it possible to reference remote pulumi stack, e,g, one in s3? StackReference seems to only reference stacks in the same backend

Getting a nasty Go error thrown when running

pulumi up

after upgrading the

pulumi

and

pulumi-aws

packages to: Pulumi: 2.3.0 pulumi (python): 2.3.0 pulumi-aws (python): 2.7.0 Looks like it's related to creating an AWS autoscaling group. Anybody else seen this/experiencing similar issues? Log here due to length: https://hastebin.com/zasaveduhu Additionally even after downgrading the python packages I get the same error (but didn't before). Do either install Go or cache Go packages somewhere or something?

am i thinking about this wrong, but i thought one could do something like

const subnet = aws.ec2.Subnet.get("some_subnet_name")

to look up an existing subnet or vpc, but apparently i need to also provide the resource id?

does pulumi delete the resource on the provider if it was imported? Is there a way one could tell?

I’m trying to use

DOCKER_BUILDKIT=1

via the

env

property on a

DockerBuild

using

awsx.ecs.Image.fromDockerBuild

. My

env

is the same as the test example ( https://github.com/pulumi/pulumi-docker/pull/95/files#diff-737a6e76d84c40fbbd438671aaf1a8c3R43 ), but I receive an error, and I don’t see the env options in there (not sure if I would though);

error: Error: ' docker build --ssh default ./app -t d2314cb1-container' failed with exit code 1

Any tips for using Docker buildkit to install private packages in a docker build? https://docs.docker.com/develop/develop-images/build_enhancements/#using-ssh-to-access-private-data-in-builds

Could not get an answer in #python channel asking here , as am not sure where to ask https://pulumi-community.slack.com/archives/CDE799L1M/p1591225877028100

I'd like to manage resources across >1 Azure subscriptions simultaneously in one Pulumi program / stack. Is there a good example for this scenario I could start from?

Can somebody help with T-shirt-sizing a following work (I'm willing to do it myself). For some weird reason which is not interesting, I need to have an option to generate standalone ARM template for pulumi stack to be used for performing actual deployment step. I'd like to re-use if possible all azure.* object model (without creating a new provider from scratch) but have something underneath translating that to ARM template instead of usual path. I do understand that ARM template is very limited in expressiveness so not everything would be even possible if this option is used.

what causes state values to become

"[secret]"

and is there a way to disable it on fields? it's causing massive diffs. I'll thread an example.

I developed on the JVM for years before picking up Haskell. I'm wondering what's the state of Java, Scala, Kotin support?

We have followed the instructions on https://www.pulumi.com/docs/guides/continuous-delivery/github-actions/ and https://www.pulumi.com/docs/guides/continuous-delivery/github-app/ but cannot get the pulumi

bot

to integrate and post to the PR. The only way we've been able to get it to post anything is to turn on the

COMMENT_ON_PR

flag (along with the

GITHUB_TOKEN

)... but that results in the ugly

pulumi --non-interactive preview

comment coming from

github-actions

bot... Are those docs up-to-date, or are we missing something critical in the setup/configuration? I've confirmed that the GitHub App is installed, and has permissions (in GH) to our repository.

hi

Is pulumi have python module to build Azure SQL Managed Instance? If none I will have to use ARM template with typescript.

Hi there, is it possible to implement own resources that participate in the Pulumi lifecycle (up/destroy) using Typescript? Are there any examples around? The use case is that I would like to output some configuration to CircleCI variables for later phases of the build process.

I'm using explicit providers only in my pulumi program, what would be the best way to disable ambient default Azure provider to reliably catch all the errors and not have something accidentally deployed in the unexpected context depending on the state of the environment where pulumi up is running? More generic question: Is there a way to inject a policy (or to do something else in a pulumi program) to completely ban usage of things like

azure.core.getClientConfig({}).then(current => current.objectId);

to enforce pulumi program to be fully hermetic, i.e. independent on the environment it runs in?

[no answer required] Just sharing feelings on the pulumification process. My cloud configuration begins to *make sense*™. I think I can even understand it now 🙂

Hi there, is it possible to start pulumi cli commands with a specific directory. I use the pulumi docker image inside my cli. In subprojects ( backend / frontend) i save the pulumi programm and stack inside an infrastructure folder and inside the cicd pipeline i want to start something like that pulumi up --path infrastructure ? Is that possible ?

Folks, what should I do if I get spurious planned deletes in the preview? I presume manually modifying the stack state via

pulumi state delete

?