pulumi-community #general

Docs

Join the conversation Join Slack

Channels

general

late-printer-99022

01/15/2020, 1:50 PM

Question - Do ingress generate Route53
A
record automatically or we need to create it?

late-printer-99022

01/15/2020, 3:29 PM

export let ingressUrl = router.status.loadBalancer.ingress[0].hostname;

I have tried pulumi.interpolate and .apply syntax but I cannot get it working. I want to get ZoneId of

url

, I got the relevant code too but it is showing error like

Argument of type 'Output<string>' is not assignable to parameter of type 'string'

😞 I have take sample code from - https://github.com/pulumi/examples/blob/master/aws-ts-static-website/index.ts

colossal-plastic-46140

01/15/2020, 3:39 PM

Hi, in pulumi is there the concept of project level configs? Or do I have to rely on creating some kind of common stack that would hold common settings?

late-printer-99022

01/15/2020, 3:41 PM

Hi Pulumi Community, I am at final round in PoC. My Ingress Code is here. https://gist.github.com/nsisodiya/d3535c78c8c95499b7efcd5805f8bda7 I want to create A record for ingress url. It involve getting ZoneId and code is faling. most of the time when I want to operate over

Output<string>

I am always stuck. Can someone help by looking at code. @flat-insurance-25294 - gist and not code dump this time 😉

flat-insurance-25294

01/15/2020, 3:42 PM

Jesus dude, you didn’t even bother to format and indent. Why are you trying to create an A record? What’s the purpose?

flat-insurance-25294

01/15/2020, 3:43 PM

What you want is an alias.

better-actor-92669

01/15/2020, 3:54 PM

Hi Pulumi community! I can give some feedback for those of you who are interested in creating a mixed GKE cluster. Even though pulumi is perfectly capable of creating a "windows" node pool (you only need to change the image type),

node_config={
    'imageType': 'WINDOWS_SAC',

windows pool is currently an early access feature and you have to send your application to GCP . Please find the link to apply below: https://cloud.google.com/blog/products/containers-kubernetes/how-to-deploy-a-windows-container-on-google-kubernetes-engine https://docs.google.com/forms/d/e/1FAIpQLSfk6PAxUqQ0k08ByqU49yy8s1sLNivHLkbQ6ltxUZf49j3C0g/viewform

fast-dinner-32080

01/15/2020, 4:28 PM

Hey Pulumi Community! I am curious if there is a Rancher 2 provider just like there is for other terraform providers today? We would like to switch to using Pulumi and I think this is the only provider I was unable to find in the docs.

powerful-football-81694

01/15/2020, 5:35 PM

Hi everyone! Is it possible to have a bi-directional dependency between two resources in Pulumi? For example: • Some resource X needs an input from an Azure App Service • The Azure App Service needs an app setting created with an output value from resource X Manually (if I were to do everything in the Azure Portal) I would do this by creating the App Service first, then resource X, then going back to the App Service to add the app setting and its value from some property of resource X. Is the same possible in Pulumi?

bitter-dentist-28132

01/15/2020, 6:43 PM

i'm doing a

pulumi.interpolate

on a string, and in the preview it showed the full string with

[secret]

where a variable would be, which was declared with

config.requireSecret

. then i decided to manually mark another interpolated variable (of type

pulumi.Output<string>

) as secret with

pulumi.secret(variable)

. that caused the entire string to become

[secret]

, so i don't get to see my config string with only the secret bits blacked out. is there a way to have that happen?

limited-rainbow-51650

01/15/2020, 6:49 PM

The

id

of a Gitlab

Group

resource is a string. In Gitlab, you can nest groups by setting the

parentId

, but this one is typed

number

. https://www.pulumi.com/docs/reference/pkg/nodejs/pulumi/gitlab/#Group-id https://www.pulumi.com/docs/reference/pkg/nodejs/pulumi/gitlab/#Group-parentId Is this something which could be aligned by regenerating the TF wrapper?

limited-rainbow-51650

01/15/2020, 6:57 PM

Meanwhile, how can I convert the

id: Output<string>

of my parent group to an

number | Promise<number> | OutputInstance<number>

which I can use as input for the

parentId

of my child group?

brave-salesmen-42327

01/15/2020, 7:51 PM

Creating a azure.compute.VirtualMachine() If I change the

osProfile: {
  customData: `#!/bin/bash\n
    echo "${someConfigValue}"`
}

It does not change the resource according to pulumi cli. Reports All Resources Unchanged, no update or destroy re-create happens. Changing something like

vmSize: "SomeOtherSize"

works fine, registers as a change and updates the vm resource.

gifted-beach-2614

01/15/2020, 9:01 PM

hey all, has anyone else had issues with high CPU usage by the python language executor while waiting for EKS clusters to be created? i’m able to reproduce it pretty easily and opened https://github.com/pulumi/pulumi-aws/issues/846 for it but am also curious if anyone else has seen this, offhand

busy-dusk-74339

01/16/2020, 12:50 AM

is anyone else having authentication issues via github?

busy-dusk-74339

01/16/2020, 12:51 AM

when i try to log in i get redirected to app.pulumi.com/github-callback/?code=… which then throws an Internal Server Error

careful-market-30508

01/16/2020, 3:22 AM

Any idea why there is no golang sdk for pulumi-kubernetes ?

late-printer-99022

01/16/2020, 4:16 AM

Hi Community, We have adopted Pulumi with lots of hope to have 100% fully Infra as code. I am pushing my company to adopt Pulumi but facing small problems here or here. Even after posting on slack, I am not getting solution. Please suggest me what to do. Example - https://github.com/pulumi/pulumi-aws/issues/848 Kindly let me know How to fix this error !!

late-printer-99022

01/16/2020, 12:01 PM

My Node.js micrservices POC is going great. finally after so much struggle, I have achieved following things. I am going to make a detailed blog once I am done with all task. 1. [done] create eks cluster 2. [done] create ECR , create services, deployments 3. [done] create Ingress for exposing services and get one ALB url per namespace. 4. [done] created A records for ALB load-balancers. http://api.syn-platform.com/ndvi/ 5. [TODO] instead of http, url must work with "https" 6. [TODO] create multiple urls on project/namespace basic Ex- https://agroproject-dev-api.syn-platform.com/userauth/ etc. 7. [TODO] http to https redirect. 8. [TODO] logging, monitoring, alerts. 9. [TODO] access to developer logging, monitoring, alerts. if anyone can review my code, it will be great. also need suggestions what else can be added in tasks list.

👍 2

late-printer-99022

01/16/2020, 7:03 PM

Target group rule is not updated. I made small change in the path of url in Ingress.

pulumi up

showed me following update.

pulumi:pulumi:Stack: (same)
    [urn=urn:pulumi:prod::syngenta-platform::pulumi:pulumi:Stack::syngenta-platform-prod]
    ~ kubernetes:<http://networking.k8s.io/v1beta1:Ingress|networking.k8s.io/v1beta1:Ingress>: (update)
        [id=syngenta-platform/syngenta-platform]
        [urn=urn:pulumi:prod::syngenta-platform::kubernetes:<http://networking.k8s.io/v1beta1:Ingress::syngenta-platform|networking.k8s.io/v1beta1:Ingress::syngenta-platform>]
        [provider=urn:pulumi:prod::syngenta-platform::eks:index:Cluster$pulumi:providers:kubernetes::syngenta-platform-provider::bb6ecf4e-2c78-4f12-8c8f-0f1b8d4ce3a7]
      ~ spec: {
          ~ rules: [
              ~ [0]: {
                      ~ http: {
                          ~ paths: [
                              ~ [1]: {
                                      ~ path: "/ngnix/*" => "/nginx/*"
                                    }
                            ]
                        }
                    }
            ]
        }

Resources:
    ~ 1 updated
    63 unchanged

Duration: 1m3s

But when I went to aws console, the path is not updated in Target Group. did I missed something?

powerful-football-81694

01/16/2020, 8:40 PM

Is there a convenient way in a CI/CD context to automate the creation of a stack if it does not already exist without throwing an error if it does already exist?

➕ 1

powerful-football-81694

01/16/2020, 8:59 PM

On a separate note: I think the task extension for Azure Pipelines has an issue with the commands it suggests in its configuration assistant UI. There is an

init

command in the dropdown, but executing the pipeline yields:

Error: unknown command "init" for "pulumi"

Did you mean this?
	new

Run 'pulumi --help' for usage.

worried-engineer-33884

01/16/2020, 9:40 PM

I'm having trouble building the pulumi nodejs sdk. Following the instructions in the README.md it seems i just need to run

make

, but it seems that you need to

npm install

first, but that doesn't work because version in package.json is set to

${VERSION}

. I don't want to hack around too much with this — is there an updated version of development spinup instructions? https://github.com/pulumi/pulumi/tree/master/sdk/nodejs#building-and-testing

flat-insurance-25294

01/16/2020, 11:33 PM

Is there any approach of getting a stack output based on what stack it is?

if (config.isProduction == false) {
    export const rdsPassword = rds.masterPassword
}

Like that wouldn’t work.

helpful-kitchen-78641

01/16/2020, 11:40 PM

I am having trouble with ACM and I suspect the cause is from using a specific region provider. I am using an S3 backend, and most of my project will take place in our default region (

us-west-2

). I am creating a certificate in

us-east-1

because I need to use it as part of a cloudfront distribution, which requires the cert to live there.

// [...]
const regionProvider = new aws.Provider('us-east-provider', {region: 'us-east-1'});

const certCertificate = new aws.acm.Certificate('cert', {
  domainName: '<http://dashboard.mydomain.com|dashboard.mydomain.com>',
  subjectAlternativeNames: [
    '*.<http://dashboard.mydomain.com|dashboard.mydomain.com>',
  ],
  validationMethod: 'DNS',
  tags,
}, {provider: regionProvider});

const certValidation = new aws.route53.Record('certValidation', {
  name: certCertificate.domainValidationOptions[0].resourceRecordName,
  records: [certCertificate.domainValidationOptions[0].resourceRecordValue],
  ttl: 300,
  type: certCertificate.domainValidationOptions[0].resourceRecordType,
  zoneId: dashboardR53Zone.id,
});

const certCertificateValidation = new aws.acm.CertificateValidation('cert', {
  certificateArn: certCertificate.arn,
  validationRecordFqdns: [
    certValidation.fqdn,
    certValidation.fqdn, // both entries use the same value/name per the AWS console
  ],
});

After running

pulumi up

I receive the following error:

Diagnostics:
  aws:acm:CertificateValidation (cert):
    error: Error describing certificate: ResourceNotFoundException: Could not find certificate arn:aws:acm:us-east-1:[REDACTED]:certificate/60c[REDACTED]df9d9.
        status code: 400, request id: 4c00[REDACTED]aef6

I can see the certificate with that ARN in the AWS console, so I'm not sure how to proceed

helpful-kitchen-78641

01/16/2020, 11:42 PM

i can move this to #aws if that's where it should go

helpful-kitchen-78641

01/17/2020, 12:04 AM

figured it out, missing the

{provider: regionProvider}

in the validation 🤦‍♂️

adventurous-vase-24423

01/17/2020, 11:19 AM

Is there a getting started guide for golang?

powerful-football-81694

01/17/2020, 1:22 PM

Hi everyone, I suspect my question went unnoticed so I’m bumping it, and clarifying a bit. I’m looking for a way to create a new stack as part of a DevOps pipeline. The stack name could for example be input using a user-configurable pipeline variable, or it could be the name of the branch that triggered the build. I want to use the Pulumi Task Extension for Azure Pipelines to initialize a new stack if it does not already exist. The obvious thing to try is to use set

command: stack

and

options: init $(StackName)

in the YAML. However, the

Pulumi@1

task has a mandatory

stack

parameter, and the task always tries to select that stack the first thing it does, before it proceeds to execute the actual command, which of course results in an error because the stack does not yet exist.

busy-dusk-74339

01/17/2020, 2:30 PM

my pulumi state has deviated from reality and i’m not sure how to fix it. i have an ec2 instance that i created from a specific commit; it has instance termination protection enabled. a colleague of mine went to add some new ebs volumes to the same stack, and when he went to run pulumi up, it wanted to replace the ec2 instance, because it thought something had changed in its security group definition (i see no such changes in the code). since instance termination was enabled, pulumi failed to delete the original ec2 instance. it did create the replacement instance, but failed before it ever got to the new s3 volumes my colleague was trying to create. even if i check out the commit that i last successfully deployed the ec2 instance from, pulumi still insists there are some sg changes. how do i get back to a sane state, here?

Title

busy-dusk-74339

01/17/2020, 2:30 PM

View count: 1