gitlab-ci: (error: Plan apply failed: deployments.apps "end-to-end-deployment-1a-og3zfpc1" not found) and 3 other deployments are failing in my kubernetes repo.

Hi folks, I'm interested in hearing about anyone who's built a true Continuous Deployment (not delivery) system to pull from code being merged to trunk, all the way to production (Kubernetes), with all the necessary accoutrements that requires. Has this been done yet?

Evaluating Pulumi as an alternative to the rather clunky Spinnaker. Been through the gamut of all other "CD" solutions and there's very very few that are really designed for Continuous Deployment, most slap the CD label on something that's really a human-gated deployment automation system

From what I see so far, this is a bit outside of Pulumi's core competency, which is more in-line with a Terraform replacement, and while it can do simple deployments to Kubernetes, complex metric gated Continuous Deployment wouldn't be the right fit. Is this fair?

I keep seeing this during rolling updates of statefulSets, it’s either picking up old data or something else entirely because the containers don’t actually exit with that status code.

~  kubernetes:apps:StatefulSet XXXXX-01310 updating [diff: ~spec]; warning: 1 Pods failed to run because: [Error] Container completed with exit code 143

The pod in question is actually just waiting to pass its

readinessCheck

I saw pulumi demo’d during TGIK, and it looked pretty cool. I want to look into it, as possibly a terraform replacement to setup a VPC, and then install k8s to that VPC. We currently use kops for k8s deployment. I’d like to move away from kops, as they lag behind k8s versions, and not ready to use aws EKS or any managed version. Are there any good resources for deploying k8s using pulumi onto ec2 instances? maybe using kubeadm?

Looking at ways to generate the JSON creds file for a GCP service account. This sample is in the getAccountKey definition:

Same variable declared twice so I’m not sure what the intention was/is

getAccountKey

does not accept the output string from the creation of the

serviceAccountKey

either

Any references to examples on using pulumi for serverless workflow orchestration/management? (Targeting AWS step functions and Azure LogicApps)

Hello. I'm just getting started with Pulumi, and I tried the JS nginx Fargate example

It booted up fine, but didn't seem to create an ingress for incoming traffic. So the url is spits out won't connect

I've just updated to 0.16.14 and now i'm getting

~   ├─ azure:keyvault:AccessPolicy                                  pulumi-keyvault-create-access                  refresh     [diff: +keyVaultId]

followed by

azure:keyvault:Secret (EventHubLeaseStorage):
    error: Preview failed: refreshing urn:pulumi:operations::operations::azure:keyvault/secret:Secret::EventHubLeaseStorage: Error checking if key vault "" for Secret "EventHubLeaseStorage" in Vault at url "<https://pulumi.vault.azure.net/>" exists: keyVaultId is empty

does pulumi have the concept of project secrets rather than stack-only secrets?

Does the circleci pulumi orb (https://pulumi.io/reference/cd-circleci.html) handle the single-queueing of pulumi runs similar to the

travisqueue

app?

sorry for x-post, but have got zero traction in the python channel, and hoping there are some more eyes here: https://pulumi-community.slack.com/archives/CDE799L1M/p1549924371007800

I'm not able to find any pulumi examples that deal with async python, and it's been a not very fun experience most times I attempt to use python

I am reading docs I can't find execution model. From the code of examples it seems the moment resource object is created all it's CRUD operations are performed, is it true? Does it mean that in a single pulumi run it is possible lets say create canary deploy and then delete it/update it (lets say scale down)? in other words, can it be an imperative program made of declarative blocks?

How come Pulumi supports so mane languages? Is it all code generated or poor Joe Duffy implement every single change N times?

Is all Joe, they have him tied to keyboards, one for each hand.

I think they said at TGIK that it’s generated

Docs:

A stack is an isolated, independently configurable instance of a Pulumi program. ... feature branches (such as feature-x-dev, jane-feature-x-dev).

Pricing:

6+ starting at $29 per stack/month

hmmm 😕

I suppose nobody is actually using them for dev branches deploy , each dev branch becomes a new component in existing stack (it probably prevents parallel deploys from happening though)

Our of curiosity is the cloud agnostic pulumi only available in typescript right now?

We have

k8s

service spun using

helm

chart and pulumi. The helm chart does not expose

loadBalancerSourceRanges

for the service definition. We would like to update the service with

loadBalancerSourceRanges

. What is the best way to do it? It is similar to

kubectl edit service

found an odd issue with AKS k8s clusters on azure.. posted an issue in pulumi-azure, but not sure if it's pulumi or the TF provider: https://github.com/pulumi/pulumi-azure/issues/182

Noticed in our pipeline - pulumi is being installed to /root/.pulumi/bin but it’s warning us there is already a copy in $PATH at /usrbin

I deleted something manually in AWS. Pulumi up expectedly shows

-   └─ aws:iam:Role                  demo-lambda-role             delete

The deletion fails, so I try to change the state.

pulumi state delete demo-lambda-role

— I get

warning: This command will edit your stack's state directly. Confirm? Yes
error: No such resource "demo-lambda-role" exists in the current state

should I not be passing in the resource name? what should I pass in?