Helm code repository-Compute Nest(Compute Nest)-阿里云帮助中心

Billing

Compute Nest is free of charge. Cloud resources provisioned during deployment incur resource fees. For details, see Cloud resource billing rule.

Prerequisites

Before you begin, make sure you have:

A GitHub repository containing Helm chart files. For this guide, fork spring-boot-chart to your own GitHub account
A Compute Nest account with access to the vendor console
GitHub repository permissions that allow Compute Nest to pull from and push to the repository

How it works

All examples use the spring-boot-chart Helm chart. The end-to-end flow is:

Connect Compute Nest to your GitHub repository.
Select Kubernetes Cluster as the environment and Helm Chart
Map ISV-defined parameters to values.yaml entries using the Ref syntax.
Choose how service instances surface their access URL: Helm Notes, a custom shell command, or a post-creation Resource Orchestration Service (ROS) template edit.
Publish the service so end users can deploy it.

Step 1: Select a method to create a service

Go to the Compute Nest console.
In the left-side navigation pane, choose My Services. On the My Services page, click the Created Services tab, then click Create Service.
On the Create Service page, select Quick Launch (From Repository). In the Repository Platform section, select GitHub.
Click Authorize Now and follow the on-screen instructions to grant Compute Nest pull and push permissions on your repository.

Skip this step if you have already granted the permissions.

Step 2: Configure service information

Configure the repository

Field	Description
Repository user/organization	Select an individual account or an organization.
Repository Name	The repository that contains your Helm chart files.
Repository Branch	The branch to import. Compute Nest auto-detects resource description files in the branch. You can proceed only after a resource description file is detected.

Configure the deployment

In the Deploy Configuration section, go to Environment Configuration and select Kubernetes Cluster. The software deployment type is automatically set to Helm Chart (the only option for Kubernetes environments). Select the chart directory — the path is relative to the root of the source repository.

Configure Helm chart values and user deployment parameters

The Helm Chart Values section lets ISVs set deployment-time values that override matching entries in the chart's values.yaml file — equivalent to helm install -f. Configure values in YAML or JSON format.

For parameters that end users must supply at deployment time (such as passwords or replica counts), define them in the user deployment parameters section and reference them in Helm Chart Values using the Ref syntax.

How parameters flow from user input to Helm

Understanding the parameter flow helps you wire inputs correctly:

An ISV defines a parameter (for example, ReplicaCount) in user deployment parameters.
The ISV references it in Helm Chart Values as Ref: ReplicaCount.
When an end user deploys the service, Compute Nest passes the user-supplied value to values.yaml before installing the chart.

The following example wires the replica count and exposes the service on port 8080 via a LoadBalancer:

YAML

replicaCount:
  Ref: ReplicaCount
service:
  port: 8080
  type: LoadBalancer

JSON

{
  "replicaCount": {
    "Ref": "ReplicaCount"
  },
  "service": {
    "port": 8080,
    "type": "LoadBalancer"
  }
}

Choose an output type

The Output setting controls how a service instance surfaces its access URL to end users. Use the following table to choose the option that matches your workflow:

Output type	How it works	Best for
Helm Notes	Displays the chart's built-in notes (from `Chart/templates/NOTES.txt`) in the service instance. End users run the kubectl commands listed in the notes to get the access URL.	Charts that already have well-written NOTES.txt files
Custom output command	ISVs write a shell script that runs in the Kubernetes cluster and prints the access URL directly.	Situations where you want users to see the URL immediately, without running any commands
Configure after service creation	Output is defined in a ROS template after the service is created.	Complex output formats, or ISVs already familiar with ROS templates

Example: custom output command for spring-boot-chart

export SERVICE_IP=$(kubectl get svc --namespace ${Namespace} spring-boot-chart -o jsonpath='{.status.loadBalancer.ingress[0].ip}')
echo "Visit URL: http://$SERVICE_IP:8080/"

After deployment completes, Compute Nest runs this command and shows the URL directly in the service instance — no extra steps for the end user.

Finish creating the service

Click Next Step: Basic Information Configuration, fill in the service details as prompted, then click Create Service.

(Optional) Edit output after service creation

This step is required only when Output is set to Configure After Service Creation.

On the My Services page, click Modify Version on the service card.
On the Modify Version page, go to User Deployment URL > Template Content and update the Outputs section of the ROS template.
Click Save Update.

Step 3: Create a service instance

On the My Services page, click the service card to open the service details page.
On the Service Details tab, click User Deployment URL in the Service Deployment section.
Specify the deployment parameters and click Next: Confirm Order. In this example, ISVs configured ReplicaCount as a user-supplied parameter, so end users specify the number of pods here. Compute Nest passes this value to the Helm Chart Values section before installing the chart.
Review the deployment summary and click Create Now.
After deployment completes, go to Service Instance > My Service Instance to view the instance. The output shown depends on the output type you configured in Step 2:
- Helm Notes — The instance displays the chart's notes. Run the listed kubectl commands to get the access URL.
- Custom output command — The access URL appears directly in the instance output.
- Configure after service creation — The custom output defined in the ROS template is shown.

What's next

Test the service — After saving the service, test it directly or pre-release it to share with specific users for validation. For details, see Test a service.
Review and publish — Submit the service for review. After Compute Nest approves it, publish the service. For details, see Review standards for publishing services and Publish a service.