pulumi-community #python

python

high-translator-22614

07/05/2019, 1:57 AM

oh, that might just be a

BucketObject

thing?

high-translator-22614

07/05/2019, 2:52 AM

given an

ResourceOptions

Resource

, can I get the provider that would be used? I want to get the current aws region for a given component resource

icy-dress-83371

07/08/2019, 7:05 PM

Anyone have any github link for Python version of Azure vm's, including load balancers, etc. Looking for an example to get started with, something like https://github.com/pulumi/pulumi-azure/tree/master/examples/loadbalancer but for python

high-translator-22614

07/08/2019, 7:25 PM

tbh, i've been using a lot of JS examples. thankfully, they largely translate

high-translator-22614

07/08/2019, 7:25 PM

the bigger problem is that pulumi-cloud and pulumi-awsx aren't available in python

high-translator-22614

07/08/2019, 7:32 PM

like, all of the existing libraries are automatically generated from terraform code? so for the most part you just have to translate the names

nutritious-microphone-86225

07/09/2019, 7:46 AM

Hi Folks, does anyone have idea on how to read value of Output in python (as string) ? E.g. I want to read assigned Ip address from an instance. As per docs this feature is not ready but maybe there is an option to use .apply in some way to read the value

nutritious-microphone-86225

07/09/2019, 7:46 AM

https://github.com/pulumi/pulumi/issues/2366

high-translator-22614

07/10/2019, 1:00 AM

await myout.future()

high-translator-22614

07/10/2019, 1:01 AM

i also like using

.apply()

as a decorator

colossal-room-15708

07/10/2019, 2:49 AM

How do I get the output of this

org

object here?

org = organizations.Organization("org", 
  feature_set="ALL"
)

roots=org.roots.apply(
    lambda roots: roots[0].id
  )

print(roots)

pulumi up

gives me:

Diagnostics:
  pulumi:pulumi:Stack (aws-scaffolding-dev):
    <pulumi.output.Output object at 0x7f599c0624e0>

high-translator-22614

07/10/2019, 2:57 AM

just hand it to whatever needs it

high-translator-22614

07/10/2019, 2:58 AM

pulumi internally resolves `Output`s and coroutines if it finds them in stuff

brief-school-94535

07/11/2019, 11:24 AM

My pulumi app just stopped working and throwing out

panic errors

when deploying a GKE cluster to GCP. My code hasn't changed and has been working for months. The errors are pointing to some golang/terraform resource issues. I've filed an issue in GitHub here: https://github.com/pulumi/pulumi/issues/2922 @white-balloon-205 any thoughts?

little-river-49422

07/11/2019, 1:51 PM

crappyform strikes again xD

nutritious-microphone-86225

07/12/2019, 9:26 AM

Hi Guys, is there any way in Python to wait until some resources are created and only after they are create to continue on with Python app ?

little-river-49422

07/12/2019, 11:31 AM

pulumi would do that on its own if one resource depends on the output of another resource

nutritious-microphone-86225

07/12/2019, 12:35 PM

right but it is between resources

high-translator-22614

07/12/2019, 3:34 PM

why? what's the use case?

high-translator-22614

07/12/2019, 3:39 PM

i mean, you could write yourself a

Gate

that takes a pile of resources and waits on them all, and then other things can depend on the gate, but that means that anything after the gate will indirectly depend on anything before the gate

nutritious-microphone-86225

07/12/2019, 4:05 PM

Any sample with example impl ? Use case - create a VM, wait for External IP allocation by GCP, then continue python script where it’s main task will be to provision VM via this external IP

high-translator-22614

07/12/2019, 4:06 PM

pulumi is all async, so you don't need to block the whole script waiting for resource creation

high-translator-22614

07/12/2019, 4:07 PM

you could pretty easily make a polling routine that runs in the background, although correctly handling dependencies in this case isn't something that's well documented

stocky-spoon-28903

07/12/2019, 4:10 PM

@nutritious-microphone-86225 If there is no natural dependency, you can specify an array of depenencies as

depends_on

in the

ResourceOptions

for the resource you want to wait

future-traffic-93788

07/12/2019, 11:45 PM

got a silly question, was playing with pulumi. How do I go about having a custom class to bake in settings for a resource

future-traffic-93788

07/12/2019, 11:45 PM

i.e.

class NginxService(Service):

    def __init__(self, service_app_name):
        app_labels = {
            "app": service_app_name
        }

        metadata = {
                       "labels": deployment.spec["template"]["metadata"]["labels"],
                   },
        spec = {
            "type": "ClusterIP",
            "ports": [{"port": 80, "target_port": 80, "protocol": "TCP"}],
            "selector": app_labels,
        }
        super(NginxService, self).__init__(service_app_name, metadata=metadata, spec=spec)

future-traffic-93788

07/12/2019, 11:45 PM

this gives a

ValueError: Unexpected type

when Pulumi is converting to protobuf

high-translator-22614

07/12/2019, 11:50 PM

I would use a factory over subclassing

future-traffic-93788

07/12/2019, 11:50 PM

is that the only approach?

high-translator-22614

07/12/2019, 11:51 PM

I assume

Service

is a pulumi-provided resource?