Stateless Application Using a Deployment
A version of the Kubernetes Stateless Application Deployment example that uses Pulumi. This example deploys a replicated Nginx server to a Kubernetes cluster, using TypeScript and no YAML.
There is an interactive Tutorial available for this example. If this is your first time using Pulumi for Kubernetes, we recommend starting there.
Pre-Requisites
Running the App
After cloning this repo, cd
into this directory and install dependencies:
npm install
Afterwards, create a new stack, a logical deployment target that we’ll deploy into:
$ pulumi stack init
Enter a stack name: k8s-nginx-dev
Now to perform the deployment, simply run pulumi up
. It will first show you a preview of what will take place.
After confirming, the deployment will take place in approximately 20 seconds:
$ pulumi up
Updating stack 'k8s-nginx-dev'
Performing changes:
Type Name Status Info
+ pulumi:pulumi:Stack k8s-nginx-k8s-nginx-dev created
+ └─ kubernetes:apps:Deployment nginx created
info: 2 changes performed:
+ 2 resources created
Update duration: 18.291517072s
This deployment is now running, and you can run commands like kubectl get pods
to see the application’s resources.
The stack’s replica count is configurable. By default, it will scale up to three instances, but we can easily change
that to five, by running the pulumi config
command followed by another pulumi up
:
$ pulumi config set replicas 5
$ pulumi up
Updating stack 'k8s-nginx-dev'
Performing changes:
Type Name Status Info
* pulumi:pulumi:Stack k8s-nginx-k8s-nginx-dev done
~ └─ kubernetes:apps:Deployment nginx updated changes: ~ spec
info: 1 change performed:
~ 1 resource updated
1 resource unchanged
Update duration: 4.324849549s
After we’re done, we can tear down all resources, including removing our stack, with a couple commands:
$ pulumi destroy --yes
$ pulumi stack rm --yes