Is there a github provider with a resource like terraforms

github_repository_webhook

?

Is it possible to detect if an action was taken in an output apply? i.e. create, update

I'm having trouble making a secondary k8s node pool for digital ocean, but nothing is standing out to me in debug mode or verbose logging to point me in the right direction. I'm getting

panic: fatal: An assertion has failed: Kubernetes GVK is: "urn:pulumi:nyc1-stage::kubernetes::digitalocean:index/kubernetesNodePool:KubernetesNodePool::kubernetes-secondary"

Has anyone got a working example, or see what I'm doing incorrectly?

Stupid question, but I can’t find the answer in the docs - for AWS functionality, am I supposed to rely on the pulumi SDK or I have to bring in the aws sdk? Asking because when I put the following in the lambda callback, the code is crashing because pulumi did not include the necessary modules into the lambda:

let secret = await aws.ssm.getParameter({
            name: "passwordParameter"
        });
console.log(secret.value);

Is it possible to write a provider that forces an update when a resource that it dependsOn is updated? I could serialize its inputs as JSON and compare the string, but I wonder if there is a more elegant way.

Could use an advice on how to attach an additional role policy to the default policy created for an implicit lambda:

aws.cloudwatch.onSchedule("test", "rate(1 minute)", async (event) => {
    try {
        let secret = await MYSSM.getValue("mySecret");
        console.log(secret);
        console.log(`current time: ${Date.now()}`);
    } catch (e) {
        console.log(e);
    }
});

In general, I’m struggling to understand on how to modify the resources that you do not create directly. Or the only way is to create and tie everything together manually?

Hello, has anyone seen this error?

error: aws:lambda/function:Function resource 'someLambda' has a problem: expected runtime to be one of [dotnetcore1.0 dotnetcore2.0 dotnetcore2.1 go1.x java8 nodejs4.3 nodejs4.3-edge nodejs6.10 nodejs8.10 nodejs10.x provided python2.7 python3.6 python3.7 ruby2.5], got nodejs12.x

I’ve also seen this error, not sure how to handle it, but its preventing me from deploying…

Hey I'm trying to deploy a nodejs app to gcloud app engine, does anyone have an example of how to do that? etc in a normal flow: 1. Pulumi creates the app engine 2. Then run npm run build 3. Navigate to folder and run gcloud app deploy

Hello, pardon my ignorance but I am going to ask a few stupid questions: I have been studying Kubernetes and Docker for the past two weeks, I am implementing a stack of services at my company that necessitated the use of containerized applications due to some of its core dependencies. To me, it seems like YAML is a poor way to go for managing infrastructure and configuration, so having a typed-language that can guide you during development in your IDE is a huge no-brainer. So Pulumi looks fantastic. One question I have though, is that even after reading the docs extensively, I am having a difficult time understanding why the message consistently seems to be for reproducibility. What I mean by that is, once you have configured and deployed your Kubernetes cluster and service stack, that is kind of it, isn't it? You wouldn't want to continuously re-deploy everything over and over by executing your infrastructure code. Another thing that I see is examples of deploying Serverless functions like Lambas, where the Lambda code is directly in the deployment function. That is neat, that eliminates the headache of having to separately manage the functions code and its deployment, but how do you develop/test that? I guess I am just unsure of what the typical workflow for this stuff is. Do you write your Pulumi infrastructure once, leave it there in your repo, and keep it around on the off-chance you have to re-deploy everything. Or just add small services every now and then? Its not like docker, where you can distribute the infrastructure file to your teammates so that they can get set up locally to develop. Sorry for the wall of text, but I am so lost in this whole Docker/Kubernetes and Serverless infrastructure, coming from a career of traditional monolith apps deployed with git-hooks to a plain cloud server.

@victorious-xylophone-55816 While it looks like what your Pulumi program would do is create everything from scratch, what it ends up doing is, it compares what you currently have to what your Pulumi program requested, and does a diff

So, if you have nothing, indeed, it will create everything from scratch - but if you already have a super similar infrastructure, Pulumi will do the minimum set of operations to get you from here to there

(This doesn't always work and sometimes Pulumi tries to recreate things that can't practically be recreated, so you have to cancel it out, but it often does!)

Ahhh, okay that IS pretty cool. So the idea still though is that generally, unless you want to programmatically deploy and manage more services, you would just

pulumi up

once in the beginning to get your stack up on your cloud provider?

Or could you use it to set up local development as well?

(I am really struggling with how to get other devs working on the current stack I have deployed on k8s in DO, it seems like some people have separate local dev configs where they use use

docker-compose

to set up the services, and then only build + push/deploy the app code Docker container to k8s, while other do Kubernetes-native development by live-sync`ing their local files to a k8s service)

(Ideally a local setup identical to the cloud would be great, but setting up k8s is a bit of a process and I am not sure if thats the way to go)

Even when you're writing your app, you can write a line,

pulumi up

, write a 2nd line,

pulumi up

and see if it did what you want, etc etc

It's actually an amazing platform for iterative development

Yeah, so, if you really do have two separate setups, a k8s setup, and a Docker-based setup, you'd write two separate Pulumi apps

How do devs usually do local development with k8s stacks?

If you have the $$, I'd instead make a super cut-down version of your k8s setup that spins the absolute minimum of nodes and have every dev create their own

That's a Good Question, I haven't done k8s stuff yet tbh, I see that as a pretty big problem tbh, having development and production being so different

Huh. Yeah, I am in a bit over my head here haha. Unfortunately I have a bit of a fire under my ass so to speak, because I have no other option than figure this out myself and I have perhaps a week or two more to do it 😅

Well thank you for taking the time to answer me, really appreciate you ❤️

I will go back to my docs reading and pound my head against things until I figure it out I suppose lmao

np!