Deployment Charts

Introduction

Devtron Apps leverage helm charts to carry out deployment of your images and configuration. Devtron includes predefined Helm charts (e.g., Deployment, Rollout, StatefulSet) that cover majority of your use cases.

For any use case not addressed by the default Helm charts, you can upload your own Helm chart and use it as a deployment chart in Devtron.

Figure 1: Deployment Charts

Tutorial

This video contains a quick walkthrough of the steps mentioned in the Preparing a Deployment Chart section of this page and the subsequent uploading of the deployment chart on Devtron.

---

Preparing a Deployment Chart

1. Create a Helm Chart

You can use the following command to create a Helm chart:

helm create my-custom-chart

Note: Chart.yaml is a metadata file that gets created when you create a helm chart. The following table consists the fields that are relevant to you in Chart.yaml.

Field	Description
Name	Name of the Helm chart (Required).
Version	This is the chart version. Update this value for each new version of the chart (Required).
Description	Give a description to your chart (Optional).

Example of Chart.yaml

Click here to view a sample 'Chart.yaml' file.

2. Create an Image Descriptor Template File

The Image Descriptor Template file is a GO template that produces a valid JSON file upon processing. It allows Devtron to dynamically inject values from the CD pipeline into your Helm chart during deployment. Therefore, details like image repository, tag, and environment are automatically populated at the placeholders specified in .image_descriptor_template.json.

• In the root directory of your chart, create a file named .image_descriptor_template.json using the following command:

  touch .image_descriptor_template.json

• Ensure the above file is created in the directory where the main Chart.yaml exists (as shown below):

  

  Figure 2: Filepath of Image Descriptor Template
• Paste the following content in .image_descriptor_template.json file:

  {
      "server": {
          "deployment": {
              "image_tag": "{{.Tag}}",
              "image": "{{.Name}}"
          }
      },
      "pipelineName": "{{.PipelineName}}",
      "releaseVersion": "{{.ReleaseVersion}}",
      "deploymentType": "{{.DeploymentType}}",
      "app": "{{.App}}",
      "env": "{{.Env}}",
      "appMetrics": {{.AppMetrics}}
  }

  You can customize this template to include only the values your deployment needs. For instance, if you only require the image repository and tag, your template would look like:

  {
      "image": {
          "repository": "{{.Name}}",
          "tag": "{{.Tag}}"
      }
  }

Got a JSON Error?

If your code editor highlights a syntax error (property or EOF error) in the above JSON, ignore it.

3. Add app-values.yaml

In the root directory of your chart, Devtron expects an app-values.yaml file. It uses this file to determine the values to be displayed on the deployment template as shown below.

Figure 3: Chart Values

The app-values.yaml file is simply a subset of your values.yaml file. Therefore, you can insert specific entries from values.yaml that you wish to display.

However, if you upload the chart without an app-values.yaml or with an empty one, your deployment template will appear blank (as shown below) or null.

Figure 4: Blank Chart Values

4. Add release-values.yaml

The release-values.yaml file contains essential values needed for deployment that aren’t covered by app-values.yaml. For example:

• Some dynamic values (such as IMAGE_TAG and IMAGE_REPO from the image descriptor JSON file) are populated here because they are needed for deployment.
• You can use autoPromotionSeconds to decide how long to keep old pods running once the latest pods of new release are available.

In the root directory of your chart, create a file named release-values.yaml with the following command:

touch release-values.yaml

Use the following content in the release-values.yaml file (edit it as per your requirement):

server:
 deployment:
   image_tag: IMAGE_TAG
   image: IMAGE_REPO
   enabled: false
dbMigrationConfig:
  enabled: false

pauseForSecondsBeforeSwitchActive: 0
waitForSecondsBeforeScalingDown: 0
autoPromotionSeconds: 30

#used for deployment algo selection
orchestrator.deploymant.algo: 1 

5. Package the chart in a tgz format

The Helm chart to be uploaded must be packaged as a versioned archive file in the format: <helm-chart-name>-x.x.x.tgz. Both <helm-chart-name> and x.x.x will be automatically fetched from the name and version fields present in the Chart.yaml file, respectively."

Note

Ensure you navigate out of the Helm chart folder before packaging it in a '.tgz' format

Run the following command to package the chart:

helm package my-custom-chart

The above command will generate a <helm-chart-name>-x.x.x.tgz file.

---

Uploading a Deployment Chart

Who Can Perform This Action?

Only super admin users can upload a deployment chart. A super admin can upload multiple versions of a chart.

Steps

• Go to Global Configurations → Deployment Charts.

  

  Figure 5: Global Configurations - Deployment Charts
• Click Upload Chart.

  

  Figure 6: Upload Chart Button
• Click Select .tgz file and upload your packaged deployment chart (in .tgz format).

  

  Figure 7: Uploading .tgz File

The system initiates the validation of your uploaded chart. You may also click Cancel upload if you wish to abort the process.

Figure 8: Cancelling Upload

Validation Checks

In the uploading process, your file will be validated against the following criteria:

• Supported archive template should be in *.tgz format.

• Chart.yaml must include the name and the version number.

• .image_descriptor_template.json file should be present.

The following are interpretations of the validation checks performed:

Validation Status	Description	User Action
Success	The files uploaded are validated (View Snapshot)	Enter a description for the chart and select Save or Cancel upload
Unsupported template	The archive file do not match the required template (View Snapshot)	Upload another chart or Cancel upload
New version detected	You are uploading a newer version of an existing chart (View Snapshot)	Enter a Description and select Save to continue uploading, or Cancel upload
Already exists	There already exists a chart with the same version (View Snapshot)	Edit the version and re-upload the same chart using Upload another chart. Upload a new chart with a new name using Upload another chart Cancel upload

---

Viewing Deployment Charts

Who Can Perform This Action?

Only super-admins can view deployment charts.

To view the list of available deployment charts, go to Global Configurations → Deployment Charts page.

Figure 9: Viewing Deployment Charts

• You can search a chart by its name, version, or description.

• You can add new charts or chart versions by clicking Upload Chart.

---

Using Deployment Chart in Application

Once you successfully upload a deployment chart, you can start using it as a deployment template for your application. Refer Deployment Template to know more.

Figure 10: Using Deployment Charts

Note

The deployment strategy for a deployment chart is fetched from the chart template and cannot be configured in the CD pipeline.

---

Editing GUI Schema of Deployment Charts

Who Can Perform This Action?

Only super-admins can edit the GUI schema of deployment charts.

Reference

This section is an extension of Customize GUI feature. Refer the document to know more about the significance of having a custom GUI schema for your deployment templates.

You can edit the GUI schema of the following deployment charts:

1. Default charts provided by Devtron (Deployment, Job & CronJob, Rollout Deployment, and StatefulSet)

2. Custom charts uploaded by you

Tutorial

Steps

In this example, we will edit the Deployment chart type provided by Devtron.

1. Click the edit button next to the chart as shown below.

   

   Figure 11: Edit GUI Schema Button
2. A GUI schema is available for you to edit in case of Devtron charts. In case of custom charts, you may have to define a GUI schema yourself. To know how to create such GUI schema, refer RJSF JSON Schema Tool.

   

   Figure 12: Editable Schema
3. You may start editing the schema by excluding existing fields/objects or including more of them. Click the Refer YAML button to view all the supported fields.

   

   Figure 13: Refer YAML Button
4. While editing the schema, you may use the Preview GUI option for a real-time preview of your changes.

   

   Figure 14: Preview GUI Button
5. Click Save Changes.

   

   Figure 15: Save Changes

Next, if you go to App Configuration → Base Configurations → Deployment Template, you will be able to see the deployment template fields (in GUI) as per your customized schema.