pulumi-community #google-cloud

damp-table-85095

09/11/2020, 5:15 AM

HI. I am trying to deploy an application to GKE. When I run I get the following errors:

Copy code

kubernetes:apps:Deployment (proxy):
    error: configured Kubernetes cluster is unreachable: unable to load schema information from the API server: Get "<https://127.0.0.1:32804/openapi/v2?timeout=32s>": dial tcp 127.0.0.1:32804: connectex: No connection could be made because the target machine actively refused it.
  pulumi:pulumi:Stack (project-dev):
    Error: invocation of gcp:container/getEngineVersions:getEngineVersions returned an error: invoking gcp:container/getEngineVersions:getEngineVersions: Cannot determine zone: set in this resource, or set provider-level zone.
        at C:\Users\Gili\Documents\project\deploy\node_modules\@pulumi\pulumi\runtime\invoke.js:136:33
        at Http2CallStream.<anonymous> (C:\Users\Gili\Documents\project\deploy\node_modules\@grpc\grpc-js\src\client.ts:155:9)
        at Http2CallStream.emit (events.js:326:22)
        at C:\Users\Gili\Documents\holdmyspot\shopping\deploy\node_modules\@grpc\grpc-js\src\call-stream.ts:186:14
        at processTicksAndRejections (internal/process/task_queues.js:75:11)

1. Why is pulumi trying to contact 127.0.0.1:32804? 2. How/where am I supposed to specify the zone associated with the cluster? I assume this is something I am supposed to pass to the k8s.Provider constructor but I don't see how. Thank you.

damp-table-85095

09/12/2020, 2:57 AM

I'm trying to put together a script that deploys onto two different zones so I am avoiding the use of the default zone configuration. Per @gentle-diamond-70147's instructions I created a

gcp.Provider

instance that specifies which zone to use but I don't understand how to tell Kubernetes which gcp.Provider to use. I mean, ultimately I want to create a single Kubernetes deployment that deploys nodes onto two different zones that I specify. How do I do that? Am I going about this the wrong way?

damp-table-85095

09/13/2020, 1:44 AM

How can one use an external file as a startup script for a new instance? Is there a mechanism to upload the file to the instance at the same time as creating it?

damp-table-85095

09/13/2020, 4:20 AM

Nevermind. I ended up using

fs.readFileSync()

to load the file into a variable.

damp-table-85095

09/13/2020, 2:20 PM

Hi. Does anyone have a workaround for this bug? https://github.com/pulumi/pulumi-gcp/issues/73

damp-table-85095

09/13/2020, 2:20 PM

I keep on having to

destroy

my entire stack (which is very slow) when debugging my startup script. Every time I update it,

pulumi up

fails with the aforementioned error.

fierce-memory-34976

09/14/2020, 9:43 AM

Hi guys, just starting with Pulumi and trying to learn something, however the documentation wasn't really useful so far. What I need is to deploy an ASP.NET Core Web API project to GCP in GKE (kubernetes). I already have a docker image of the service in GCP but I'm unsure on how to deploy the app and set the proper ingress service through Pulumi. Can anyone help or does anyone know a guide/tutorial where I could read more on this?

fierce-memory-34976

09/14/2020, 11:20 AM

I'm trying to run pulumi on ubuntu (WSL) and I get the following error: error: configured Kubernetes cluster is unreachable: unable to load schema information from the API server:... My current kube config file is configured for localhost (docker desktop with k8s) but I would like it to work with GCP.. How do I get it to work with GCP without creating a cluster there first and getting that configuration (it worked like that) because pulumi should be creating the configuration..

fierce-memory-34976

09/14/2020, 11:21 AM

do I need to create a service account with needed configuration and use that for gcp auth (json key) or what does this error actually mean?

fierce-memory-34976

09/15/2020, 9:15 AM

Hi, getting the following warnings when trying to deploy a K8S service and deployment to GCP GKE:

warning: configured Kubernetes cluster is unreachable: unable to load schema information from the API server: Get "https://<redacted>/openapi/v2?timeout=32s": x509: certificate signed by unknown authority

Anyone knows what should I do to get rid of this?

fierce-memory-34976

09/15/2020, 12:56 PM

Another question.. I need to setup load balancing with SSL on GKE but I'm unsure what to do.. Anyone has any example of this with Pulumi?

fierce-memory-34976

09/15/2020, 4:14 PM

Is it possible to create this with Pulumi? https://cloud.google.com/kubernetes-engine/docs/how-to/managed-certs#setting_up_the_managed_certificate

red-area-47037

09/15/2020, 9:37 PM

Is there a way to directly spin up a GKE cluster on the RAPID release channel with the most current version (1.18.6-gke.3504)? So I dont have to upgrade the master nodes and then the worker nodes?

fierce-memory-34976

09/17/2020, 4:47 PM

anyone having issues with refused connection (pulumi->gcp) at the moment? i'm getting strange errors

Error while waiting to delete default node pool: Error waiting for removing default n

ode pool: error while retrieving operation: Get "<https://container.googleapis.com/v1beta1/projects/><redacted>/locations/us-east4/operations/operation-1600360932143-8939a57c?alt=j

son&prettyPrint=false": dial tcp [2a00:1450:400f:809::200a]:443: connect: connection refused

chilly-garage-80867

09/18/2020, 12:30 AM

howdy, I am trying to deploy a k8s cluster on gcp

chilly-garage-80867

09/18/2020, 12:30 AM

I created a ConfigFile in order to install some CRDs

chilly-garage-80867

09/18/2020, 12:30 AM

I am getting error: Duplicate resource URN

chilly-garage-80867

09/18/2020, 12:30 AM

I am pretty sure it’s unique though

chilly-garage-80867

09/18/2020, 12:32 AM

Copy code

::kubernetes:yaml:ConfigFile$kubernetes:<http://apiextensions.k8s.io/v1beta1:CustomResourceDefinition::images.caching.internal.knative.dev';|apiextensions.k8s.io/v1beta1:CustomResourceDefinition::images.caching.internal.knative.dev';> try giving it a unique name

chilly-garage-80867

09/18/2020, 3:28 AM

I am also seeing

Copy code

Exception: invocation of kubernetes:yaml:decode returned an error: error converting YAML to JSON: yaml: line 128: mapping values are not allowed in this context

chilly-garage-80867

09/18/2020, 3:28 AM

if I actually go and manually apply this yaml with kubectl, it’s fine. yamllint.com says it’s valid too

damp-elephant-82829

09/18/2020, 6:44 PM

How interesting

damp-elephant-82829

09/18/2020, 6:44 PM

Step #12 - “Pulumi Create Stack”: opts=opts, Step #12 - “Pulumi Create Stack”: File “/usr/local/lib/python3.7/site-packages/pulumi_gcp/storage/get_project_service_account.py”, line 103, in get_project_service_account Step #12 - “Pulumi Create Stack”: email_address=ret.email_address, Step #12 - “Pulumi Create Stack”: AttributeError: ‘dict’ object has no attribute ‘email_address’ Step #12 - “Pulumi Create Stack”: error: an unhandled error occurred: Program exited with non-zero exit code: 1

damp-elephant-82829

09/18/2020, 6:44 PM

get_project_service_account on the storage module is not working anymore?

damp-table-85095

09/20/2020, 2:27 AM

How can one associate specific SSH keys to a GCP instance at creation time?

fierce-memory-34976

09/23/2020, 10:25 AM

Hi, i have created a new stack with:

pulumi stack init dev --secrets-provider="gcpkms://<redacted>"

and in my code I have:

var config = new Config();

and later in code in an env var definition:

Value = config.RequireSecret("dev-db-connection-string")

and when i try pulumi up i still get:

Missing Required configuration variable 'projectName:dev-db-connection-string'

`please set a value using the command `pulumi config set projectName:dev-db-connection-string <value>`` the secret in GCP is called obviously

dev-db-connection-string

what am I missing?

damp-elephant-82829

09/25/2020, 10:10 AM

I0925 084254.872241 61 langruntime_plugin.go:171] langhost[python].RunPlan(pwd=/deployment,program=.,...,dryrun=true) success: progerr=Program exited with non-zero exit code: 1 I0925 084254.873243 61 source_eval.go:178] EvalSourceIterator ended with an error: an unhandled error occurred: Program exited with non-zero exit code: 1 I0925 084254.873475 61 plan_executor.go:219] planExecutor.Execute(...): incoming event (nil? true, &{0xc000aee160}) I0925 084254.873674 61 eventsink.go:86] eventSink::Error(<{%reset%}>an unhandled error occurred: Program exited with non-zero exit code: 1<{%reset%}>) I0925 084254.873913 61 step_executor.go:321] StepExecutor worker(-1): StepExecutor.waitForCompletion(): waiting for worker threads to exit I0925 084254.874094 61 step_executor.go:321] StepExecutor worker(-2): worker exiting due to cancellation I0925 084254.874283 61 step_executor.go:321] StepExecutor worker(-1): StepExecutor.waitForCompletion(): worker threads all exited I0925 084254.874504 61 plan_executor.go:254] planExecutor.Execute(...): step executor has completed I0925 084254.874868 61 plan_executor.go:131] planExecutor.Execute(...): exiting provider canceller + pulumipulumiStack barchart-import-master create error: an unhandled error occurred: Program exited with non-zero exit code: 1 I0925 084254.882010 61 ignore.go:44] Explicitly ignoring and discarding error: rpc error: code = Canceled desc = grpc: the client connection is closing + pulumipulumiStack barchart-import-master create 3 errors

damp-elephant-82829

09/25/2020, 10:10 AM

opts=opts, File “/usr/local/lib/python3.7/site-packages/pulumi_gcp/storage/get_project_service_account.py”, line 96, in get_project_service_account email_address=ret.email_address, AttributeError: ‘dict’ object has no attribute ‘email_address’ <{%reset%}>)

damp-elephant-82829

09/25/2020, 10:10 AM

I am still stucked here, dict has no attribute email address

clean-helmet-84712

09/25/2020, 6:33 PM

First - thanks for making Pulumi, I've been teaching myself it all week and it's a pleasure to write real IaC. Second - is it possible to get the current GCP project that's being operated on, regardless if it's set by GOOGLE_PROJECT, GOOGLE_APPLICATION_CREDENTIALS?