pulumi-community #general

boundless-airport-99052

11/08/2019, 11:25 AM

To give more details about the usecase: When I update the

max_prepared_transactions

configuration of an Azure PostgresSQL server, I have to restart the database. Neither pulumi nor terraform seems to be able to restart the database. So I want to know if this resource (

azure.postgresql.Configuration

) has been created during the

pulumi up

and if true, trigger a manual restart outside of pulumi.

better-actor-92669

11/08/2019, 12:09 PM

Hi! Is there a way to change regional quotas with pulumi?

Copy code

ERROR: (gcloud.container.clusters.create) ResponseError: code=403, message=Insufficient regional quota to satisfy request for resource: "IN_USE_ADDRESSES". The request requires '9.0' and is short '1.0'. The regional quota is '8.0' with '8.0' available.

Open the GCP Console and navigate to IAM & admin -> Quotas.
Filter the quotas by selecting your region under Location.
Check the box next to Compute Engine API CPUs, then click EDIT QUOTAS.
Follow the steps to increase the quota. Quotas are not immediately increased.

better-rainbow-14549

11/08/2019, 12:26 PM

i'm having a weird issue with TS 3.7.2 and pulumi.all - it seems to think all the parameters are the same type. is this known / is there a workaround please?

curved-ghost-20494

11/08/2019, 1:28 PM

Hi All, I'm looking to "componentize" or "template" some of my basic Pulumi setup (still learning)". I'd like to be able to do something like

npm install @mynpm/infra-templates

and then use `createStaticSite('``www.foo.com``', './projects/foo-dot-com/www')` to abstract away the setup for creating an S3 based static site.

curved-ghost-20494

11/08/2019, 1:29 PM

Does anyone know what the Pulumi vocab is for describing this? I've seen stack mentioned? I think I can just do it in JS, but then I wouldn't be able to access the exported stack outputs?

colossal-plastic-46140

11/08/2019, 2:13 PM

Hi Pulumi Folks, Is it possible for the stack files to live in a different directory in which the `__`_`main`_`__.py` lives?

better-rainbow-14549

11/08/2019, 2:13 PM

yeah e.g.

pulumi login file://$PWD

better-rainbow-14549

11/08/2019, 2:14 PM

oh stack files i read state files, ignore me

elegant-crayon-4967

11/08/2019, 8:31 PM

I’m spinning up an ec2 backed ECS cluster and the ASG creates instances fine and all is good, but the ECS cluster never assigns the instances to it. Does anyone have a good code example of spinning up a simple ec2 backed AWS ECS cluster? I’m using the command

ecsCluster.addAutoScalingGroup(awsxEcsClusterASG)

and thought that would be the connection to tie the 2

gorgeous-postman-12842

11/08/2019, 8:56 PM

new to pulumi here. I tried creating a new azure-typescript project. Deployed it. Doing

pulumi stack ls

is always empty. I have logged in using

--local

gorgeous-postman-12842

11/08/2019, 8:57 PM

Copy code

❯ pulumi stack ls
NAME  LAST UPDATE  RESOURCE COUNT

billions-lock-73409

11/08/2019, 9:03 PM

Any thoughts on a multi account strategy using a single Pulumi code base? In my use case we have a non-prod AWS account and a prod account. My initial thought is to just create stacks within the Pulumi project that map to the separate accounts, has anyone done this or something similar (or different)?

gorgeous-postman-12842

11/08/2019, 9:06 PM

@billions-lock-73409 if my understanding is correct, stacks don't store information about accounts. In my case (Azure), before invoking pulumi, I must be logged in and setup to a specific Azure subscription. Pulumi is agnostic of the fact that I can have multiple accounts and multiple subscriptions.

billions-lock-73409

11/08/2019, 9:10 PM

Yeah that makes sense, we use Circle for automation of the deployments, but when performing a local deployment we use aws-okta to SSO into the respective accounts. I just wasn’t sure if anyone had any specific workflows with dealing for separate accounts. One example is we need a few resources to exist in the non prod account (s3 buckets) and not in prod, so we’ve just created a specific section where those resources are only included during build if the stack matches a specific environment.

gorgeous-postman-12842

11/08/2019, 9:17 PM

@billions-lock-73409 a better approach to avoid complication would be to script it. In my situation we have automated our ARM Template deployments using Powershell scripts which take custom parameters and ensure the deployment is done to the provided environment in addition to some other custom logic.

billions-lock-73409

11/08/2019, 9:18 PM

Okay, I see. So your deployments are push-button style and not automated? (we have a pause in place after the preview runs on circle)

gorgeous-postman-12842

11/08/2019, 9:18 PM

The issue I mentioned about

pulumi stack ls

yielding empty result, I have commented on an existing bug here: https://github.com/pulumi/pulumi/issues/2679#issuecomment-551992056 Possibly it should be a new bug.

gorgeous-postman-12842

11/08/2019, 9:21 PM

@billions-lock-73409 Sadly, ARM deployments for us have to be done manually using the powershell scripts I mentioned. ARM doesn't provide a

destroy

functionality so one needs to be very cautious while using it. Also, our infra deployments are not done daily. We have decent setup of ARM and corresponding template files but by design ARM leaves a lot to be wanted.

billions-lock-73409

11/08/2019, 9:22 PM

Gotcha

billions-lock-73409

11/08/2019, 9:22 PM

Thanks for the feedback 👍

important-leather-28796

11/08/2019, 10:29 PM

Why does an unchanged gke/k8

ClusterRoleBinding

show as an update?

bumpy-breakfast-57632

11/08/2019, 10:59 PM

did the way pulumi gets aws credentials change somehow? because now it fails when it didn't before, I have them in the Pulumi.stack.yaml file and it seems to ignore it and use whatever's in the .aws folder, but it was working previously without any .aws folder

early-intern-90238

11/09/2019, 1:20 AM

Pretty sure RabbitMQ is melting my brain....

early-intern-90238

11/09/2019, 1:21 AM

I am working on getting this working with Pulumi, its pretty straightforward...I'll add it the examples when I am done. https://github.com/rabbitmq/rabbitmq-peer-discovery-k8s/tree/master/examples/k8s_statefulsets

fancy-translator-86814

11/09/2019, 11:55 AM

Any update on CircleCI orb for Pulumi? It seems great new features are coming in new version such as

stack_init

stack_rm

, and

stack_output

. I really looking forward to it.

better-actor-92669

11/11/2019, 3:00 PM

Hi guys! If a pulumi resource hasn't been created properly and I want to force the redeploy. How can I do that? For example I couldn't create a GKE cluster, because of

Cluster "gke-cluster-1" has terminal state "ERROR" with message "Retry budget exhausted (10 attempts): Google Compute Engine: A route in one of the active peer networks hides the address space of the new subnetwork."

Now I want to recreate it, but I get

Plan apply failed: googleapi: Error 409: Already exists: projects/.../zones/europe-west3-c/clusters/gke-cluster-1., alreadyExists

I removed this resource from pulumi state directly with

pulumi state rm urn

, but the behaviour is the same.

delightful-lizard-41466

11/11/2019, 3:35 PM

Hi! Is there a proper way to reuse an "azure.storage.Account" resource created by a previous Pulumi stack? I'm trying to use the StackReference but without success, suggestions ? Thanks

creamy-shoe-96075

11/11/2019, 5:03 PM

Hi everyone, Im looking at app.pulumi.com and looks like pricing is related to amount of stacks. In this case if I do prefer “high granularity”(it has a lot of benefits) of stacks, app.pulumi.com becomes really pricy. What’s you opinion in this matter guys?

modern-army-90330

11/11/2019, 7:54 PM

Has anyone run into issues with outputs on failed updates?

enough-ability-20396

11/12/2019, 5:02 AM

Hi! I inherited a GCP project setup done via Pulumi. When I run

pulum preview

pulumi up

commands, I see the following error log. Any ideas on what might be going on?

Copy code

$ pulumi preview
Previewing update (solvvy/<redacted>):

     Type                 Name                       Plan     Info
     pulumi:pulumi:Stack  <redacted>           1   error

Diagnostics:
  pulumi:pulumi:Stack (solvvy-ml-main):
    error: Running program '/Users/amey/work/code/infrastructure-live/gcp-projects/<redacted>/project' failed with an unhandled exception:
    Error: invocation of gcp:organizations/getActiveFolder:getActiveFolder returned an error: invoking gcp:organizations/getActiveFolder:getActiveFolder: Folder not found
        at monitor.invoke (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/@pulumi/pulumi/runtime/invoke.js:74:33)
        at Object.onReceiveStatus (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/grpc/src/client_interceptors.js:1205:9)
        at InterceptingListener._callNext (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/grpc/src/client_interceptors.js:568:42)
        at InterceptingListener.onReceiveStatus (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/grpc/src/client_interceptors.js:618:8)
        at callback (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/grpc/src/client_interceptors.js:845:24)
        at Function.module.exports.loopWhile (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/deasync/index.js:71:23)
        at promiseResult (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/@pulumi/pulumi/utils.js:70:13)
        at Object.liftProperties (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/@pulumi/pulumi/utils.js:102:19)
        at Object.getActiveFolder (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/node_modules/@pulumi/organizations/getActiveFolder.ts:37:25)
        at Object.<anonymous> (/Users/amey/work/code/infrastructure-live/gcp-projects/solvvy-ml/project/core-project.ts:9:21)

Permalink: <https://app.pulumi.com/solvvy/><redacted>/<redacted>/previews/a98b9966-82b6-4be3-9cf3-4d1b3d3b937c

FWIW I have

owner

access to the underlying GCP project and I am able to run

pulumi

commands on other stacks in the same project.