I am having a difficulty installing a plugin: pulumi plugin install resource postgresql v0.18.0 [resource plugin postgresql-0.18.0] installing Downloading plugin: 14.94 MiB / 14.94 MiB 100.00% 53s error: installing [resource plugin postgresql-0.18.0] from : moving plugin: rename C:\Users\grandm1\.pulumi\plugins\resource-postgresql-v0.18.0.tmp543610203 C\Users\grandm1\.pulumi\plugins\resource postgresql v0.18.0 Access is denied.

So here's an architectural question I've got with Pulumi, and I'd love some thoughts on it. When you use AWS API Gateway, your calls to the AWS API make stateful changes to the gateway's configuration. However, it also has the concept of a "deployment", which has to be bumped in order to actually push those changes live. Because Pulumi and not AWS's state is the canonical record for me, I would like a deployment to be bumped each and every time I make a change to a method, an integration--whatever. Is there a good way to express this in Pulumi? I'm struggling to think of one.

I'm interested in testing out HAProxy in my k8s cluster. They recently released an official HAProxy Ingress controller but there does not appear to be a helm chart for it. Is it possible to work with & customize k8s yaml files in Pulumi? i.e. I want to apply the HAProxy yaml as described here: https://www.haproxy.com/documentation/hapee/1-9r1/traffic-management/kubernetes-ingress-controller/#install-the-haproxy-ingress-controller and customize it as described here https://www.haproxy.com/documentation/hapee/1-9r1/traffic-management/kubernetes-ingress-controller/#customizing-the-haproxy-ingress-controller.

Is there a flag I am unaware of with

pulumi up

that will perform all deletes before creates? This would be really useful in some scenarios. Our current one is moving from manifests to helm. The resource names are the same, but the URNs are different. Having pulumi delete before create would mean we don’t need to resolve these conflicts manually.

My

pulumi up

becomes very slow sometimes. It shows

Previewing update

and nothing else for a couple minutes. That goes on for some time, and then everything seems fast again. Last time Node process was high on CPU, but now everything seems idle. I'm on Windows. Anyone experiencing this?

Seeing some weird behavior on refreshes, ec2 security-group tag keys not being read as strings

Hi hackers! Do pulumi provides a kind of api gateway + lambdas offline testing commands?

Does

pulumi.awsx.autoscaling.AutoScalingGroup

only support ECS? If so, why? Seems like it could support instances that don't use the ECS AMI. https://github.com/pulumi/pulumi-awsx/blob/a8dd50c8f5be5137686553cfb4a4d62041481f1b/nodejs/awsx/autoscaling/launchConfiguration.ts#L279-L414

Hi Pulumi! I'm working on making a StackDriver log export to BigQuery. I'm running into an error where my (unique) log writer is unable to write to the dataset, even though I'm giving it a role that should allow it to do so. Do the BigQuery Datasets also require explicit permissions for that role, or am I setting the role-binding incorrectly?

Here's my code

Is there any Pulumi magic that would let me Update a dataset I just created with new accesses in the same deployment? That would enable me to pass the "uniqueWriterIdentity" to the dataset, but I need to make the dataset first before I know the unique writer identity. As a workaround I'm using Google's generic writer identity, but they now recommend against that practice.

Hi everyone. When you save a stack to pulumi.com it saves it as <org><project><stackname> but locally or using a bucket, it just saves stacks in the same location and you can't have duplicate stack names in different projects. I've seen some github issues around this. Is this a known issue or is there a workaround(except creating unique stacknames)?

Is there a better way to do:

const awsIamPolicy = new aws.iam.Policy('rsin-provider', {
  name: `${config.environment}-rsin-provider`,
  description: 'Allow rsin-provider to connect to ElasticSearch',
  policy: JSON.stringify(pulumi.interpolate`{
      Version: '2012-10-17',
      Statement: [
        {
          Action: ['es:*'],
          Effect: 'Allow',
          Resource: "${elasticSearchDomain.arn}/*"
        },
        {
          Effect: 'Allow',
          Action: ['dynamodb:DescribeTable', 'dynamodb:Query', 'dynamodb:Scan'],
          Resource: ${clientsTable.arn}
        },
        {
          Effect: 'Allow',
          Action: ['dynamodb:DescribeTable', 'dynamodb:Query', 'dynamodb:Scan'],
          Resource: ${auditLogsTable.arn}
        },
        {
          Effect: 'Allow',
          Action: ['dynamodb:DescribeTable', 'dynamodb:Query', 'dynamodb:Scan'],
          Resource: ${linksTable.arn}
        },
        {
          Effect: 'Allow',
          Action: ['dynamodb:DescribeTable', 'dynamodb:Query', 'dynamodb:Scan'],
          Resource: ${notificationsTable.arn}
        },
        {
          Effect: 'Allow',
          Action: ['dynamodb:DescribeTable', 'dynamodb:Query', 'dynamodb:Scan'],
          Resource: ${usersTable.arn}
        }
      ]
    }`)
});

Is there a good way to pass

pulumi config

values into a Cloud Function's handler code without including the pulumi package in it? I was having memory issues when I included it.

Ok, so I used dynamic resources to make some setup on Droplet from DigitalOcean in create method of my custom provider, is there any way to easily check in my dynamic resource if that droplet was updated or is newly created?

hi folks, i’m trying to stand up an EKS cluster per some of the ‘getting started’ documentation for crosswalk. it seems some of the out-of-the-box k8s services are not starting due to no nodes being ready. my nodes are reporting ‘not ready’ due to the network not being initialized:

$ kubectl describe nodes | grep 'KubeletNotReady' | head -n1
   Ready            False   Fri, 21 Jun 2019 13:57:30 -0700   Fri, 21 Jun 2019 13:50:29 -0700   KubeletNotReady              runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:docker: network plugin is not ready: cni config uninitialized

to my eye, it seems to be same/similar issue as: https://github.com/pulumi/pulumi-kubernetes/issues/578 https://github.com/pulumi/pulumi-eks/issues/174 is there a known workaround?

Looking at the aws cli reference for

create-data-source

there’s an option for HTTP to have

authorizationConfig

option, but i don’t see that in

@pulumi/aws/appsync/dataSource.d.ts

. Any ideas how to add authorization to an AppSync HTTP request?

In this example of aws API Gateway:

How do you set the Lambda integration as a Proxy?

Ah looks like it defaults to it nevermind sorry

Hmmm with the awsx.apigateway.API class, how do I make the method response an http Proxy?

Like this:

It defaults to no integration response like this:

Thats the code I have and the eventHandler: blah is a Lambda that points to a src folder

Ah that's what it is I think, the Lambda body has base64 encoded json in it, i wanted it to return just a string or json.. I think it's because of binary media type / in api gateway

Hi, guys. What do you think about connecting domain and ssl in pulumi script ? or it's better do this things manually ? I try to do it via this example - https://github.com/pulumi/examples/tree/master/aws-ts-static-website But try to change it to container and Fargate service.

In the awsx.apigateway.API class, is it possible to drill down into apigateway to set it to a 'regional' distribution instead of edge-optimized?

Is it possible to move a project from one pulumi account to another?