Let's introduce you to the Ocuroot client by building and deploying a set of interrelated services to a local Docker instance.
Through deploying these services, you'll see how Ocuroot manages missing dependencies, adding and populating new environments and deleting environments.
Ordinarily, you may prefer to manage local containers with docker compose, but in this example, we'll be deploying
everything separately to illustrate how Ocuroot would operate in your datacenter or cloud environments.
You will need the following installed on your local machine:
We'll be working on a local copy of the quickstart repo, so you'll need to clone it.
git clone https://github.com/ocuroot/quickstart.git
cd quickstart
Before we can deploy our services, we need somewhere to deploy them to. We need an environment.
For the quickstart, we have a config file that will do that for you, environments.ocu.star.
Kick off a release from that file to create a staging environment:
ocuroot release new environments.ocu.star
To get us started, this only creates a staging environment. You can see the staging environment config by viewing the state:
$ ocuroot state get @/environment/staging
{
"attributes": {
"frontend_port": "8080",
"type": "staging"
},
"name": "staging"
}
For the purposes of this quickstart, state is stored locally in the .store directory.
Now we have an environment, we can release something to it. We'll start with the "frontend" service.
ocuroot release new frontend/package.ocu.star
This should output something like this:
✓ build (30.399s)
Outputs
└── quickstart/-/frontend/package.ocu.star/@1/call/build#output/image
└── quickstart-frontend:latest
› deploy to staging
Pending Inputs
└── quickstart/-/network/package.ocu.star/@/deploy/staging#output/network_name
The staging deployment has a pending input from network/package.ocu.star. This is
because we need a shared docker network for the staging environment.
To satisfy the dependency for the frontend, we need to release the network.
ocuroot release new network/package.ocu.star --cascade
This will both create the network needed for the frontend, and then continue releasing the
frontend now that it is unblocked. The latter behavior is enabled by the --cascade flag.
The frontend should now be running and you can view it at http://localhost:8080. You'll see three errors about unreachable services, this is because we need to deploy them!
Run these three commands to deploy the backend services.
ocuroot release new time-service/package.ocu.star
ocuroot release new weather-service/package.ocu.star
ocuroot release new message-service/package.ocu.star
Once complete, go back to the frontend and you'll see messages from these services.
We now have a fully populated staging environment! Ocuroot also includes a web UI to view your state. You can start it by running:
ocuroot state view
This will start a local server showing the contents of your state store. Once started, open a browser to http://localhost:3000 and take a look around.
Now we have our staging environment fully populated and visually tested, we can set up production!
Add the following to environments.ocu.star:
register_environment(environment(
name="production",
attributes={
"type": "production",
"frontend_port": "8081",
}
))
Then release these changes and deploy all services.
ocuroot release new environments.ocu.star --cascade
This will create the new production environment, and deploy all your release to it in the correct order.
Once this is complete, you'll be able to load the production frontend at http://localhost:8081, there should be a line on the page indicating that the environment is "production".
You'll now have a bunch of containers running in your local Docker. View the list by running:
docker ps -f name=^quickstart- --format "{{.Names}}"
You'll see containers for both production and staging.
Let's clean up after ourselves, first off, we'll delete our production environment. First, let's delete the intent.
ocuroot state delete @/environment/production
This will update your intent for the production environment, but won't actually perform any actions.
In a typical deployment of Ocuroot, this change to intent will automatically trigger ocuroot work cascade
on your CI platform.
Now we can run ocuroot work cascade to actually perform the deletion, but first, let's look at what it's about
to do with the --dryrun flag.
ocuroot work cascade --dryrun
This outputs something like:
[
{
"ref": "@/environment/staging",
"work_type": "delete"
}
]
Which indicates that we're going to delete the realized state of the staging environment, including all resources.
So now we can execute all work required to remove this environment with a single command.
ocuroot work cascade
If you run the docker ps command above again, you'll only see the staging containers.
See if you can adapt the above commands to delete the staging environment as well.
You can also look at the state with ocuroot state view and you will see that there are
no environments listed, although there is a full record of all historical deployments.
This was just a taste of what you can do with Ocuroot, and there's plenty more to explore even within this repo! Feel free to have a look around to see how everything's configured. You could also try: