pulumi-community #python

damp-elephant-82829

09/11/2020, 8:20 PM

Hello guys, I have a failure when trying to import inside the “main.py” files from modules but if I run python from the same folder where I run pulumi up and I try to import my module it works

damp-elephant-82829

09/11/2020, 8:20 PM

Untitled

astonishing-quill-88807

09/14/2020, 2:00 PM

I'm having some issues with trying to render a stack export that is combining values from multiple resources. I have a composite resource where I was able to export a dictionary of some of its attributes. Now, I'm trying to add information about other resources created in the same stack, but when I preview the output it's showing that a bunch of existing data will be deleted, with some other data being added in a seemingly random fashion. My assumption is that it's due to some indeterminacy of how the async processes resolve. Has anyone dealt with this before or have advice? I tried using the

Output.all(...).apply(...)

approach but instead got errors from deep in the Pulumi runtime about dicts not being callable...

average-school-38756

09/16/2020, 7:32 PM

Anyone have a working VS Code debugging implementation for code run by the CLI? Hoping someone can spot the issue in mine. Inspired by https://github.com/pulumi/pulumi/issues/1372#issuecomment-583086422 i set up a

.vscode/launch.json

looking like this:

Copy code

{
  "version": "0.2.0",
  "configurations": [
    {
      "name": "Python: Attach using Process ID",
      "type": "python",
      "request": "attach",
      "processId": "${command:pickProcess}"
    }
  ]
}

i put this code at the beginning of my

___main___.py

, with a breakpoint as a sanity check (and i have a VS Code breakpoint on the same line):

Copy code

import debugpy
debugpy.listen(5678)
debugpy.wait_for_client()
debugpy.breakpoint()

i run

pulumi up

which shows this output (appears as though

wait_for_client

is indeed working):

Copy code

$ pulumi up
Previewing update (dev):

And i hit F5 to trigger the process attachment. i choose the process called "pulumi up" in the dialog. After ten seconds or so, i get a prompt in VS Code saying "Timed out waiting for debug server to connect."

curved-waitress-45592

09/18/2020, 1:12 AM

hi there, is there a way to Pulumi/Python without the venv mechanism that's built in? we use anaconda a lot and would love to stick to it and not have to juggle multiple environments during our development.

agreeable-toddler-25853

09/18/2020, 7:34 AM

HI All, Is there a way to get a list of all the IAM policies

worried-plastic-45846

09/22/2020, 8:51 AM

Hello, Since this morning, I'm getting this error, on a project that worked until now:

Copy code

aws:codebuild:Project (develop-Codebuild-Project-build):
    error: aws:codebuild/project:Project resource 'develop-Codebuild-Project-build' has a problem: Attribute must be a whole number, got 480
    error: aws:codebuild/project:Project resource 'develop-Codebuild-Project-build' has a problem: Attribute must be a whole number, got 60

How do I know which attribute is causing an issue? Using -v does not provide a single error message, only debug. In the entirety of the output, the number 60 and 480 do not appear at all.

curved-cartoon-90054

09/22/2020, 10:16 PM

hi I am trying pulumi crossguard for first time i am getting below error error: no Pulumi.yaml project file found

Copy code

steps:
mkdir policypack && cd policypack
pulumi policy new aws-python
pulumi preview --policy-pack <path-to-policy-pack-directory>

clever-plumber-29709

09/23/2020, 5:11 AM

Is there a way to make https://github.com/joeduffy/aws-tags-example/blob/master/autotag-py/taggable.py someting redistributable? I dont like the ide of havinf to copy it on each resource i want to auto tag

helpful-processor-86468

09/24/2020, 11:13 AM

error: Exception calling application: code() takes at most 15 arguments (16 given)

- this is what I'm getting when using dynamic resource when trying to do

pulumi destroy

. It was working fine for some time and now it's broken without any code change. What is happening?

curved-cartoon-90054

09/24/2020, 7:19 PM

hi can we use pulumi crossguard as general purpose policy/decision making engine similar to opa? Like decisions not related to deployment eg: in opa we can write rules like validating password where user can provide their password as data to opa and get a response if the password is matching the rules

average-school-38756

09/25/2020, 9:24 PM

Any reason a policy violation wouldn't be printed to the

pulumi preview

output? i ran

pulumi preview --policy-pack my-pack --policy-pack-config my-config.json

, i see the pack named under "Policy Packs run", and i know

report_violation("my message")

was executed because i stepped through it in the debugger. This is with a custom Python pack i started writing. Running against an AWSGuard TypeScript policy does show policy violations.

shy-football-10348

09/30/2020, 2:59 AM

Does anyone have an example of enabling ALB access logs? I have the following code which doesn’t work:

Copy code

alb_http_bucket_key = aws.kms.Key(
    "jupyterhub-alb-http-bucket-key",
    deletion_window_in_days=10,
    description="This key is used to encrypt bucket objects",
)

alb_http_bucket = aws.s3.Bucket(
    "jupyterhub-alb-http-logs",
    server_side_encryption_configuration={
        "rule": {
            "applyServerSideEncryptionByDefault": {
                "kms_master_key_id": alb_http_bucket_key.arn,
                "sseAlgorithm": "aws:kms",
            },
        },
    },
)

alb = aws.lb.LoadBalancer(
    "jupyterhub-alb",
    security_groups=[sgroup.id], 
    subnets=jupyterhub_priv_subnets,
    internal=True,
    access_logs=aws.lb.LoadBalancerAccessLogsArgs(
        bucket=alb_http_bucket.bucket_domain_name,
        enabled=True
    )
)

The bit I cannot get right is the

access_logs=

field. I directly referenced another function that has an example using the same data type that

access_logs

requires and I still cant get it to work 😞

shy-football-10348

09/30/2020, 2:59 AM

appreciate any help! thanks in advance.

agreeable-toddler-25853

09/30/2020, 3:00 AM

Hi, Any ideas on how we can upload a directory instead of a file to a s3 bucket using pulumi

nutritious-shampoo-16116

09/30/2020, 3:41 PM

anybody has a good way to catch errors that happen in futures? We hit sometimes this problem, having resources deleted cause errors are swallowed silently if any happens using

.apply

nutritious-shampoo-16116

09/30/2020, 3:42 PM

it makes also debugging pretty hard

polite-pillow-78450

09/30/2020, 4:26 PM

In a new stack, I am trying to fetch the Fargate ECS cluster with:

import pulumi_aws as aws

cluster = aws.ecs.get_cluster(cluster_name=‘fargate’)

polite-pillow-78450

09/30/2020, 4:26 PM

I get the following error:

Exception: invocation of awsecs/getClustergetCluster returned an error: required configuration keys were missing

polite-pillow-78450

09/30/2020, 4:26 PM

I haven’t managed to find any issues in the Pulumi provider or terraform repos.

polite-pillow-78450

09/30/2020, 4:27 PM

Latest version of

pulumi

and

pulumi-aws

polite-pillow-78450

09/30/2020, 4:27 PM

Any ideas?

faint-table-42725

09/30/2020, 6:43 PM

This is happening when you try to access

result.settings

? Do you mind sharing a snippet of your code?

polite-pillow-78450

09/30/2020, 6:44 PM

No, this is literally just the call to

get_cluster

polite-pillow-78450

09/30/2020, 6:44 PM

I am not accessing anything from the result.

polite-pillow-78450

09/30/2020, 6:44 PM

import pulumi_aws as aws

cluster = aws.ecs.get_cluster(cluster_name=‘fargate’)

polite-pillow-78450

09/30/2020, 6:44 PM

Everything else is commented out.

polite-pillow-78450

09/30/2020, 6:46 PM

$ aws ecs list-clusters

{

“clusterArns”: [

“arnawsecsus east 1XXXXXX:cluster/fargate”,

]

}

polite-pillow-78450

09/30/2020, 6:47 PM

describe-clusters

polite-pillow-78450

09/30/2020, 6:50 PM

updated to latest pulumi library (not CLI) fixed the issue.