Pulumi Cloud Framework only supports AWS currently, correct? ( this is what the public webpage says) If so, is there an ETA on GCP and Azure support? Ref: https://pulumi.io/quickstart/cloudfx/index.html

What is the standard approach to running tests of lambdas for example? If they are stored in the same repository there isn't any way of invoking just that lambda through Pulumi, so do people tend to launch a stack and run tests against that? Or break out the function logic to perform unit tests? I'm trying to understand the development lifecycle and what other people are doing

Just a heads up, I am occasionally seeing some oddities with instances of

aws.s3.BucketObject

- Pulumi reports that an object has been created, and it is present in the state and website UI, but not actually present in AWS. Doing a

pulumi refresh

causes the object’s non-existent status to be correctly reflected in the state, and then another

pulumi up

creates it correctly. I’d give you a better repro case if I could, but it’s not happening consistently, so hard to pin down.

Looks like the year on the changelog needs to be updated to 2019 for the last 3 releases

this is killing me and I'm not sure of the issue

googleapi: Error 400: Invalid Service Account email ([object Object])., badRequest

passing a service account into a function to create IAM membership based on the email of said serviceAccount

I do an export of the created account and email address looks fine

Hi all, we just started evaluating Pulumi. We currently have the problem that our AWS deployment file exceeds the maximum allowed size because unnecessary large files in node modules are included. So is there a way to exclude parts of node modules ?

Looking for some TypeScript pro tips? Rumor has it that @white-balloon-205 is good at this sort of thing. Join us at 11am PST today, and ask any questions live, or here in advance https://twitter.com/PulumiCorp/status/1093187930661769222

Live in 3 minutes ^^

was there an update to the kubernetes provider? I’m seeing

create-replace

for all of my k8s resources and the only diff is adding the default namespace

--kubernetes:core/v1:Service: (delete-replaced)
                [id=default/external-digibyte-61651]
                [urn=urn:pulumi:cointainers-staging::cointainers::standard:Deploy$kubernetes:apps/v1:StatefulSet$kubernetes:core/v1:Service::external-digibyte-61651]
            +-kubernetes:core/v1:Service: (replace)
                [id=default/external-digibyte-61651]
                [urn=urn:pulumi:cointainers-staging::cointainers::standard:Deploy$kubernetes:apps/v1:StatefulSet$kubernetes:core/v1:Service::external-digibyte-61651]
              ~ metadata  : {
                  + namespace  : "default"
                }

How do I write an asynchronous Azure Function App (async await) in Pulumi?

Is there a way to disable checking for replica counts on k8s statefulSets? Some of our services take 12+ hours to become ready

Diagnostics:
  kubernetes:apps:StatefulSet (litecoin-0163):
    error: Preview failed: 2 errors occurred:

    * Resource 'litecoin-0163' was created but failed to initialize
    * Failed to observe the expected number of ready replicas

I’m seeing some

read

actions on recoded resources, any idea what those are?

Hey everyone. Trying to deploy one of our platforms described by a

Helm

chart into pulumi. We use the

aws-alb-ingress-controller

which means we have the following line in our ingress.

paths:
          - path: /*
            backend:
              serviceName: ssl-redirect
              servicePort: use-annotation

This sets up automatic ssl forward at ALB level in AWS. However Pulumi doesn’t understand this third party rule and is not fully deploying showing:

kubernetes:extensions:Ingress  core-django               creating...  No matching service found for ingress rule: "***-staging.***.com/*" -> "ssl-redirect"

Kubernetes question on a local Mac machine. Is minikube the same as the Kubernetes cluster that now ships with Docker’s app? I mean, can I use Docker’s Kubernetes cluster as a drop-in replacement for minikube? I’m debugging a deployment error (a timeout when creating a

k8s.core.v1.Service()

) and wondering if there’s a problem linked to the

isMinikube

setting.

Another question on Azure App Function (still have first question open): Deployment of Azure Function App did not seem to copy NPM dependency as well (says it can't find it). Does Pulumi not support this or do I need to do something special to ensure the 3rd party dependency is bundled along with my code?

anybody see an issue where

pulumi preview

fails due to missing plugin. You then run

pulumi plugin install

and it doesn't resolve any?

npm install ended grabbing a new version of the kubernetes plugin just now but still no gcp

Running into this error

gcp:dns:RecordSet (frontend):
    error: unexpected null property rrdatas[0]

Just a feature request: it would be nice if the filter arguments to

aws.ec2.getInstance

(and indeed all the

get*

functions) could take Inputs as well as literal values. I can work around it by calling it inside an

apply

, but it seems like it would be pretty straightforward to do that by default and keep client code cleaner.

I've written a project that will create and manage some ECR instances in multiple regions. Someone already manually created some of them, but I'd like my pulumi stack to manage those instances now. How can I import existing things into my stack?

Hi guys. I just joined. I'm evaluating pulumi for our company and try to get some stuff up that we definitely need. What I am stuck on at the moment is appsync - does pulumi have full support, like everything to create functions, pipeline resolvers, basically the lastest appsync capabilities? Also does it support the RDS Serverless Aurora DataSource ?

FYI, I am using aws.lambda.Function() and I set timeout: 5 — instead of setting the lambda timeout to 5 minutes, it gets set to 5 seconds

Any use Pulumi to setup an AWS Batch service? Looking for examples, thanks! https://aws.amazon.com/batch/

What’s the best way to package up a lambda from code in the same directory? I like the ability to create lambdas in-line with pulumi, but I’m needing a bit more structure to the lambda and more JS files and would prefer to not have to zip the code up each time

can I create a lambda just from a folder path? and specify extraIncludePackages?