(I am hoping the answer to friel above might answer this @white-balloon-205) To limit access, I’d like to have multiple stacks. One stack might provision resources like s3 buckets, and the other might mainly be “serverless” functions. In creating resources, I could usually expect that my resource would always be called

my-bucket

, but Pulumi assigns unique IDs to each resource, so it becomes

my-bucket<some_id>

. Am I right in assuming that if I want to access a second-stack resource, I either have to create some registry of resource names or manually pass them to the second stack?

We'll be hosting our first meetup at our Seattle HQ on Feb 21st - join us if you can! Also, if you'd like to host a meetup in your city - let me know (DM or marc@pulumi.com) https://www.meetup.com/Pulumi-Seattle/events/258138687/

Is there a place I can look for some python examples with Pulumi? I see a lot of typescript

Office hours is back tomorrow. Join @white-balloon-205 to talk about the latest and greatest features and issues. https://twitter.com/PulumiCorp/status/1090301558418477056

Question , how do I convert an

Output<string>

to

string

? One of the function returns an

Output<string>

and the other one expects a

string

.

I added an

aws.cloudwatch.Dashboard

to my code and received this after running

pulumi up

. No idea what it means. These are my package deps:

"@pulumi/pulumi": "0.16.9",
"@pulumi/aws": "0.16.5",
"@pulumi/cloud-aws": "0.16.2",
"@pulumi/aws-infra": "0.16.3-dev.1548354465"

error:

error: TypeError: message.getPropertydependenciesMap is not a function
        at Function.proto.pulumirpc.RegisterResourceRequest.serializeBinaryToWriter (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/@pulumi/aws/node_modules/@pulumi/pulumi/proto/resource_pb.js:775:15)
        at proto.pulumirpc.RegisterResourceRequest.serializeBinary (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/@pulumi/aws-infra/node_modules/@pulumi/pulumi/proto/resource_pb.js:696:43)
        at serialize_pulumirpc_RegisterResourceRequest (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/@pulumi/aws-infra/node_modules/@pulumi/pulumi/proto/resource_grpc_pb.js:95:26)
        at Object.final_requester.sendMessage (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/grpc/src/client_interceptors.js:806:37)
        at InterceptingCall._callNext (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/grpc/src/client_interceptors.js:419:43)
        at InterceptingCall.sendMessage (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/grpc/src/client_interceptors.js:464:8)
        at InterceptingCall._callNext (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/grpc/src/client_interceptors.js:428:12)
        at InterceptingCall.sendMessage (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/grpc/src/client_interceptors.js:464:8)
        at ServiceClient.Client.makeUnaryRequest (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/grpc/src/client.js:536:21)
        at ServiceClient.method_func [as registerResource] (/home/kenny/compute_software/demo-applications/basic-web/pulumi/node_modules/grpc/src/client.js:950:43)

anybody have a good Windows and Mac/Linux workflow for AWS role assumption? Several tools but most have no Windows support

trying some out but figured I'd see if the community here used anything they liked

Hi is there plan to support digitalocean? Currently documentation doesn't say so.

Hi guys! Anyone have already used slots in an azure webapp or functionapp with Pulumi? I'm wondering how to manage them

Is there currently a bug (or something else I don’t understand) with the “uncommited changes were present at the time of this update” flag on the Pulumi website? I’m seeing this even when

git status

reports “nothing to commit, working tree clean” both before and after I run

pulumi update

.

Pulumi now supports Atlassian identity - @clever-sunset-76585 has the details https://blog.pulumi.com/pulumi-now-supports-atlassian-identity

I am trying to get the service load balancer that is created within

k8s

? I have deployed this using helm chart. I am able to get the hang of the service withsomething like this

istio.getResource("v1/Service", "istio-ingressgateway");

which returns the

k8s

service object https://github.com/pulumi/pulumi-kubernetes/blob/d301148551488034535fc2a66c29aa5a2f86380b/sdk/nodejs/provider.ts#L8435 From this service how do I get the reference to load balancer that was created?

hi guys, is the

aws.elasticloadbalancingv2.getLoadBalancer

working with tags filter? When I'm using the name or the arn, I get back the load balancer, but with tags it says

Search returned 7 results, please revise so only one is returned

. It has a unique name, so that could not be the problem. What is strange to me, that if it based on terraform

aws_lb

data source, then why the tags argument not listed here? https://www.terraform.io/docs/providers/aws/d/lb.html

We are now live on - ask @white-balloon-205 anything!

^^ Currently working through GKE setup

anyone get a MongoDB Atlas provider to work in Pulumi? It's basically the only thing currently keeping me from migrating from Terraform to Pulumi

hello - im trying to use the

namespace

param in

ChartOps

to deploy a chart from a public repo into a specific name space but it ended up in default rather than the specified name space. has anyone had any luck with this?

I'm new to Pulumi and trying to create a lambda to handle an Alexa skill. Is there an Alexa skill event that I can directly add, or is there a way to generically define an event?

I'd like to use Pulumi to prototype a workflow system using SQS queues. In a lambda that reads off of one queue, I would like to write an SQS message to another queue. How can I write to an SQS queue from within a lambda? Note: not interested in using a step function.

From

k8sIngress

, How do I get actual LoadBalancer in this case the

ELB

?

I’m trying to figure out how to get AWS AppSync working. It seems like I can create the service itself, but I can’t actually define the GraphQL schema or any of the resolvers or mapping templates anywhere. Am I missing something? Or is this just using the very limited Terraform implementation behind the scenes?

Is there a one-liner for linux installs to install a specific version of the CLI?

I want to set up automation and lock to a specific version.

One of my stacks is stuck in deleting a object which I looked does not exists in

aws

. How do I by pass that?

I'm working on locking access to map IAM roles to specific namespaces within my EKS cluster. The role mapping will change frequently, whereas the cluster infra will not. E.g. A single EKS cluster will serve as a sandbox for dozens of teams. We will create and destroy namespaces often as teams reserve some of the cluster for testing or deployment. Each team will be granted an IAM Role that will map via the

aws-auth

configmap to a Kubernetes user that is limited to actions within a specific namespace. Hence, the aws-auth configmap will be updated often and quickly move out of sync with the initial definition for it in when we

new eks.Cluster

. Can you recommend a Pulumi way to manage the namespaces, IAM roles, kubernetes roles, rolebindings & usernames?

"error: Preview failed: unknown stack" - pipeline was working last week

Constant, reproducible error on

pulumi up

Are there more complicated examples in node/ts of the dynamic provider being used somewhere? I seem to keep running into issues around how the code is captured/executed. Once I write my code in a way that passes serialization and preview, it will often error when applying because something wasn't captured during serialization (sometimes my variables, sometimes ts things like __awaiter). I'm curious why serializing the code is necessary for the dynamic provider to function? Wouldn't it always use the assets one has locally? I'll also make some issues and minimal repros for the serialization issues I've noticed sometime this weekend.